Effects of Electrodeposition Mode and Deposition Cycle on the Electrochemical Performance of MnO2-NiO Composite Electrodes for High-Energy-Density Supercapacitors

Nanostructured network-like MnO2-NiO composite electrodes were electrodeposited onto stainless steel substrates via different electrodeposition modes, such as chronopotentiometry, chronoamperometry, and cyclic voltammetry, and then subjected to heat treatment at 300°C for metal oxide conversion. X-ray diffraction, field emission scanning electron microscopy, and transmission electron microscopy were used to study the crystalline natures and morphologies of the deposited films. The electrochemical properties were investigated using cyclic voltammetry and charge/discharge tests. The results revealed that the electrochemical performance of the as-obtained composite electrodes depended on the electrodeposition mode. The electrochemical properties of MnO2-NiO composite electrodes prepared using cyclic voltammetry exhibited the highest capacitance values and were most influenced by the deposition cycle number. The optimum specific capacitance was 3509 Fg−1 with energy and power densities of 1322 Wh kg−1 and 110.5 kW kg−1, respectively, at a current density of 20 Ag−1 in a mixed KOH/K3Fe(CN)6 electrolyte.


Introduction
Interest in supercapacitors has increased recently because of the high demand for energy production and storage resulting from rapid population growth and high energy consumption in the industrial sector [1]. This area of research not only faces the challenge of maintaining renewable and sustainable energy sources but must also find a method to store this energy effectively [2][3]. In general, such energy can be stored in three ways: electrical, chemical, and electrochemical energy storage. In chemical storage, the energy takes the form of potential energy that can be released by chemical reactions. When chemical energy is converted into electricity, this method is described as electrochemical energy storage and is commonly used in This study investigated the effects of different electrodeposition modes on the capacitance performance of NiO-MnO 2 composite electrodes in Na 2 SO 4 electrolyte. To the best of our knowledge, no comparative study of composite binary metal oxides prepared by different electrodeposition modes has been reported previously, except for one report on single-metal oxide electrodes. We show that all the studied MnO 2 -NiO samples contained homogeneous distributed porous structures. CV measurements revealed that potentiodynamically deposited MnO 2 -NiO electrodes had high specific capacitance, which can be attributed to the presence of pseudocapacitive MnO 2 nano-particles connected to the electrically conductive NiO. Finally, we demonstrated that the morphological structure, deposit thickness, and electrochemical performance were influenced by the number of deposition cycles. The capacitance decreased as the number of deposition cycles increased because of the ineffective wettability properties at the electrode/electrolyte interfaces and the lower intercalation/deintercalation activity into or from the electrodes. The optimum electrode was studied in three different electrolytes, i.e., Na 2 SO 4 , KOH, and mixed KOH/ K 3 Fe(CN) 6 . We previously addressed MnO 2 -NiO electrodes deposited via CP with different Ni concentrations in the deposition solution; please consult that work for the experimental details [14].

Materials and Methods
All the chemicals used in this study were analytical grade and were used without further purification. The deposition electrolyte was prepared by mixing 0.01-M Mn acetate tetrahydrate (Mn(CH 3 COO) 2 .4H 2 O), 0.25-M Ni acetate tetrahydrate (Ni(CH 3 COO) 2 .4H 2 O), and 0.8-M sulfuric acid (H 2 SO 4 ) to obtain a homogeneous solution. The Mn-Ni oxide (MnO 2 -NiO) films were deposited onto 2×2-cm 2 stainless steel (SS) substrates using different deposition modes. The deposition process was conducted in three electrode systems with a carbon rod, Ag/AgCl, and SS as the counter, reference, and working electrodes, respectively. The three electrodeposition modes studied here were CP, CA, and CV. To deposit the Mn-Ni hydroxide using the CP mode, the current was maintained at 8 mA for 10 min of deposition time (sample named CP), whereas a constant voltage of 1.5 V was applied to deposit the Mn-Ni hydroxide via CA (sample named CA). In the CV mode, the electrodeposition was conducted in the voltage range from 0 to 1 V at a scan rate of 20 mVs −1 with 7 deposition cycle (sample named CY7). Finally, the deposited samples were rinsed with distilled water before being subjected to post-heating at 300°C for 6 h. The CV mode (CY7) generated the best results, and thus, the effect of the number of deposition cycles in CV mode was studied further. To this end, the numbers of deposition cycles tested were 4 cycles, 10 cycles, and 13 cycles, and the corresponding electrodes were named CY4, CY10, and CY13, respectively.
The powder X-ray diffraction (XRD) patterns were recorded using a D8 Advance X-Ray diffractometer-Bruker AXS instrument with CuK α monochromatic radiation at 40 kV. The elemental composition was studied using energy-dispersive X-ray (EDX) spectroscopy (Oxford Instruments). Field emission scanning electron microscopy (FESEM) and transmission electron microscopy (TEM) images of the electrodes were captured using Jeol JSM-7600F and Jeol JEM 2100F instruments. The amount of material loaded onto the SS substrate was weighed using a microbalance (Sartorius CPA225D). The electrochemical measurements were performed using three-electrode methods and an Autolab PGSTAT30 potentiostat/galvanostat. Platinum wire was used as a counter electrode, and Ag/AgCl was used as the reference electrode. CVs were recorded in a potential window ranging from −1 to +1 V versus Ag/AgCl with a scan rate of 5 mVs −1 , whereas galvanostatic cycling was recorded in a potential window of −0.6 V to 1 V with a current density of 1 Ag −1 . Electrochemical impedance spectroscopy (EIS) was conducted in the frequency range of 10 mHz to 100 kHz. The specific capacitance (C) can be calculated from both the CV and charge/discharge (CD) results using Eqs 1 and 2, respectively [18]: where I is the oxidation/reduction current, dt is the time differential, m is the mass of the active material, and ΔV is the operating potential.
where I is the discharge current, dV dt is the change of the discharge potential with the discharge time, and m is the mass of the active materials.
Additionally, the energy (E) and power (P) densities can be expressed as Eqs 3 and 4, respectively [19]: where V is the operating potential, and R s is the equivalent inner resistance of the device.

Results and Discussion
The crystal structure of Mn-Ni oxide was examined by XRD. The XRD patterns of Mn-Ni oxide powder deposited by CP, CA, and CY7 are shown in Fig  diffraction planes of NiO [4,[20][21]. The synthesized CY7 electrode had relatively weak peak intensities, suggesting that the deposited MnO 2 -NiO particles are in a poorly crystalline state, which could contribute to these particles having a larger specific area than those in a highly crystalline film [21]. No change in the chemical compositions of MnO 2 and NiO was noted between the electrodes deposited using different modes. The EDX results of the deposited MnO 2 -NiO powder is shown in Fig 1B. The compositions and weights (%) of Ni, Mn, and O are displayed and confirmed the successful deposition of MnO 2 -NiO on the SS substrates. The possible mechanisms underlying the deposition binary metal oxides via different electrodeposition modes can be described by the equations presented below. When the potential is applied to the deposition cell, H 2 O is reduced on the working electrode, as described by Eq 5.
Water reduction produces OH − ions, which are believed to play a vital role in the deposition of metal oxides. The condensation of Mn + and Ni + ions occurs where the OH − ions are localized. The binding affinity between Mn + /Ni + cations and OH − leads to the nucleation of metal hydroxide particles. Metal oxide particles then come to rest on the substrate during the annealing process after the hydroxide is efficiently oxidized in the complex, as described by Eqs 6 and 7 [14,22].
The nucleation and growth of the oxide particles may be affected by the electrodeposition mode, resulting in alterations of the metal oxide particles' morphologies, but the elemental properties are not affected, as demonstrated by XRD.
The as-prepared Mn-Ni oxides obtained via CA, CP, and CY7 had the same crystal structures but different morphologies, which were investigated by FESEM and TEM, as shown in Fig 2. The FESEM images revealed that all of the MnO 2 -NiO samples contained homogeneously distributed porous structures. Fewer particle agglomerations were found in the CY7 electrode sample (Fig 2C). Closure inspection of the TEM images revealed that the obtained CA and CY7 electrodes displayed smaller pore sizes (~2-4 nm) than the CP electrode (~10-15 nm). These results indicate that the CA and CY7 samples are mesoporous materials, which allow electrolyte ions to easily diffuse through the electrode bulk and, thus, enhance the number of faradic reactions that occur during electrochemical operation [9][10].
The morphologies of the electrodes produced using different numbers of deposition cycles was investigated using FESEM (Fig 3). This figure shows that varying the number of deposition cycles resulted in different morphologies. After the fewest number of deposition cycles tested (CY4 electrode, Fig 3A), clear interconnected metal oxide particles with a porous structure were observed; however, the flake structure was not able to grow because of the limited number of cycles. Fig 3B-3D) show the formation of a clear flake structure and reveal that it becomes increasingly flakes as the number of deposition cycles increases. These compact/dense flakes increase the deposited mass load. The possible mechanisms at work during the deposition process are explained below. The initial formation of stable interconnected MnO 2 -NiO is believed to occur early during the deposition (around the forth cycle) because of the instantaneous nucleation process. Subsequently, a progressive nucleation process takes place, in which the flake structure grows on the top of the previously formed nuclei. As the number of deposition cycles increases, the growth rate of the progressive nucleation also increases, resulting in denser flakes and compact grains [23][24].
The TEM image of a CY4 electrode is shown in Fig 3E. This figure reveals no flake structures in the deposits and more highly distributed pores than CY7 (Fig 2C). The elemental composition of the MnO 2 -NiO deposits in the CY4 electrode was analysed using high-resolution TEM lattice images (Fig 3E), (inset)). These images confirmed that the measured interplanar spacing corresponds to MnO 2 and NiO. Furthermore, the periodic lattice fringe of 0.31 nm corresponds to the (310) plane of crystalline α-MnO 2 , whereas the lattice spacing of 0.24 nm corresponds to the (111) plane of NiO. These interplanar spacings are in agreement with the XRD results of MnO 2 -NiO [25][26]. To determine the supercapacitive behaviours of the studied electrodes, CV was performed in 0.5-M Na 2 SO 4 in the potential window from −1 to 1 V at a scan rate of 1 mVs −1 and are displayed in Fig 4. A pair of symmetric anodic and cathodic peaks are clearly evident in each CV profile, implying the existence of two reversible faradic reactions in the tested samples. Regarding charge storage in MnO 2 electrodes, two mechanisms have been suggested, as described in Eqs 8 and 9 [27][28]: Based on the reaction above, the OH− ion plays an important role in NiO charge storage. Indeed, decreasing the OH − concentration in the Na 2 SO 4 aqueous electrolyte solution might lead to less-significant contributions from NiO redox reactions; NiO mainly enhances the capacitance that corresponds to EDL storage [21,18]. The largest encircled area was exhibited by the CY7 electrode, indicating that this material has a higher electrochemical activity and capacitance. The calculated specific capacitance values for the CP, CA, and CY7 electrodes at a scan rate of 1 mVs −1 were 435, 458, and 500 Fg −1 , respectively. The high specific capacitance of the CY7 electrode may be attributable to its lower agglomeration, well-distributed mesoporous structure, and low crystallinity (amorphousness), as discussed in the XRD, FESEM, and TEM sections. Therefore, the phase structure and morphology, in addition to the porous nanostructure, of MnO-NiO significantly influence its electrochemical properties.
To obtain additional information on the electrochemical behaviour of the synthesized CY7 electrode obtained using different numbers of deposition cycles, the CV curves of the CY4, CY7, CY10, and CY13 electrodes in 0.5-M Na 2 SO 4 electrolyte were collected and are shown in Fig 5A. The peak current decreased as the number of deposition cycles increased, suggesting that the diffusion of cations into the MnO 2 -NiO electrode decreased. The denser, more compact flake structure may have prevented cations from migrating into the electrode material, thereby decreasing Na + or OH − ion adsorption. Fig 5B shows the CD curve of all electrodes obtained with different numbers of deposition cycles at a current density of 1 Ag −1 and a voltage range from −0.6 V to 1 V. The non-linearity of the curve is attributable to the faradic redox reaction that occurs in this voltage range [14,29]. The CY4 electrode exhibited a longer discharge time, indicating that its charge-storage behaviour is more efficient than those of other electrodes. The specific capacitances of CY4, CY7, CY10, and CY13, based on the mass of deposited MnO-NiO calculated from a CDC curve, were 769 Fg −1 , 557 Fg −1 , 556 Fg −1 , and 227 Fg −1 , respectively. The trend of the specific capacitance was inversely proportional to the mass of the electrode [30], as shown in Fig 5C. The observed decreasing trend of the capacitance relative to the mass is attributable to the increasing resistance, which influences the ion conduction, as shown in the Nyquist plot in Fig 6. The ion transport resistances of all samples was investigated by EIS. The equivalent circuit based on the Nyquist plots (Fig 6A) was fitted using Nova software, and the parameters are shown in Table 1. R s was determined from the x-intercept of the arc in the high-frequency region, which reflects the ohmic resistance of the electrolyte, the contact resistance, and the internal resistance of the material. Because R ct was measured in the semicircle region, it corresponds to the transfer resistance of the CD reaction at the electrolyte/electrode interface [31]. The R s and R ct values of the CY4 electrode were smaller than those of the CY7, CY10, and CY13 electrodes, as revealed in Table 1 (Fig 6A). This finding indicates good wettability of the electrode/electrolyte interfaces and suggests that the cation intercalation/deintercalation activity into or from the CY4 electrode is easier than those of the other electrodes [14]. Additionally, the CY4 electrode's smaller R s values indicate that it has lower resistivity, which may be the result of the increased surface area of the deposited composite electrode structure [32]. Thus, the CY4 electrode has higher specific capacitance and exhibits better electrochemical performance than the CY7, CY10, and CY13 electrodes.
To obtain additional more information on the processes that contribute to the electrode's capacitance, the Bode plots presented in Fig 6B and 6C can be divided into three frequency regions: high-(f>10 Hz), medium-, and low-frequency regions (f<1 Hz). Typical capacitive  behaviour was observed in both plots throughout the whole frequency range. In the low-frequency region, all deposited films presented a slope of~−1 in the f-|Z| plots. The phase angles in the f-ɸ plots were between −70°and −55°. The intermediate values of the impedance magnitude |Z| and phase angle (ɸ) were obtained in the medium-frequency region, whereas in the high-frequency region, the capacitance was nearly zero because the |Z| is independent. Additionally, the phase angle began to decrease to almost zero as the frequency was further increased. The frequency at a phase angle = −45°in the f-ɸ plots (Fig 6C) is known as the frequency response region, and higher frequencies at ɸ = −45°indicate better capacitive responses. The highest frequency response found here was 2.22 Hz for the CY4 electrode, and this value was related to this material's reasonably fast electron/proton transport response, which led to high capacitance [33].
After confirming that the CY4 electrode was the optimum electrode in this system, its performance was further investigated in different alkaline electrolytes. The CVs of the CY4 electrode in 0.5-M Na 2 SO 4 , 0.5-M KOH, 0.04-M K 3 Fe(CN) 6 , and mixed 0.5-M KOH/0.04-M K 3 Fe (CN) 6 electrolytes were obtained sweeping the potential from −0.5 V to 0.5 V at a scan rate of 5 mVs −1 and are shown in Fig 7. A well-defined pair of anodic peaks (A 0 ) and a cathodic peak (C 0 ) are observed cantered at potentials of approximately +0.20 and −0.05 V (vs. Ag/AgCl), respectively, in Na 2 SO 4 electrolyte (Fig 7A). These peaks can be attributed to the redox reaction Effects of Electrodeposition Mode of MnO 2 -NiO Composite Electrode for Supercapacitors dominated by MnO 2 , according to Eqs 8 and 9 (presented above). The KOH electrolyte was chosen because it has a higher OH − concentration, which plays an important role in NiO reactions during charging/discharging (Eq 9) [34]. The combined contributions of MnO 2 -NiO in the KOH electrolyte increased the current response, as shown in Fig 7B. The high current response of this system results from the complete utilization of NiO and the relatively small K + cation radius (3.31 Å) compared to that of Na + ions (3.35 Å), which facilitates the passage of K + ions into the electrode matrix during the charging process [35].
When 0.04-M K 3 Fe(CN) 6 was added to the 0.5-M KOH electrolyte, an additional pair of anodic peaks at approximately +0.25 V (A 02 ) and a cathodic peak at +0.17 V (C 02 ) arose ( Fig  7C). These peaks were attributed to the redox reaction of K 4 Fe(CN) 6 to K 3 Fe(CN) 6 , which is consistent with the CV plot obtained using K 3 Fe(CN) 6 electrolyte (Fig 7D) [36]. The drastic increase in the current response observed in this system occurred because of the appearance of two significant faradic reactions originating from the redox couple [Fe(CN) 6 ] 3− /[Fe(CN) 6 ] 4− in the electrolyte (Eq 10) and the redox reactions in the highly electroactive electrodes (Eq 11).
Redox electrolyte : Redox electrode : where M is the Ni 2+ or Mn 2+ cation, and 1 n z. The detailed mechanism was presented in a previous study [37]. When the electrode is charged, the active materials are oxidized, and the electrolyte [Fe(CN) 6 ] 3− accepts an electron via the reduction of hexacyanoferrate (III) to (II); the hexacyanoferrate ions act as "electron carriers" (Fig 8A). When the electrode is discharged, the reaction is reversed: The hexacyanoferrate ions act as "electron donors", which means that [Fe(CN) 6 ] 4− is converted to [Fe(CN) 6 ] 3 − , and the released electrons reduce the electrode active materials (Fig 8B). This helps to improve the capacitive performance of the active material [37][38]. The specific capacitances calculated from the CV at 5 mVs −1 are as follows: 474 Fg −1 , 780 Fg −1 , and 5130 Fg −1 for Na 2 SO 4 , KOH, and mixed KOH/K 3 Fe(CN) 6 , respectively. The importance of K 3 Fe(CN) 6 was confirmed by the enhancement of the specific capacitance.
The CD profiles of the CY4 electrode in Na 2 SO 4 , KOH, and mixed KOH/ K 3 Fe(CN) 6 electrolytes at a current density of 20 Ag −1 are shown in Fig 9A. The specific capacitances calculated from the discharge curves were 46 Fg −1 , 583 Fg −1 , and 3509 Fg −1 in Na 2 SO 4 , KOH, and mixed KOH/K 3 Fe(CN) 6 , respectively. A high specific capacitance was also obtained in mixed KOH/K 3 Fe(CN) 6, which is consistent with the CV results. The CD of CY4 in mixed KOH/ K 3 Fe(CN) 6 electrolyte was further investigated using different applied current densities and is plotted in Fig 9B. The obviously non-linear curve, which includes two potential plateaus at approximately −0.14 V and 0.18 V, was inferred to result from the redox reaction of KOH and K 3 Fe(CN) 6 [29]. The energy and power densities of CY4 in mixed KOH/ K 3 Fe(CN) 6 were 1322 Wh kg −1 and 110.5 kW kg −1 , respectively, at a current density of 20 Ag −1 .
The cycling performance of the CY4 electrode in three different electrolytes is shown in Fig  9D. The capacitance retention exhibited a drastic decrease during continuous charging-discharging. The specific capacitance retention values in Na 2 SO 4 , KOH, and mixed KOH/K 3 Fe   Fig 9. (A) CD profiles of the CY4 electrode's current density at 20 Ag −1 in three different electrolytes, (B) CD profiles collected using different applied current densities in mixed KOH/K 3 Fe(CN) 6 , and (C) specific capacitance retention after the 1500 th cycle at a scan rate of 10 mVs −1 in the potential range from −0.5 V to 0.5 V.
doi:10.1371/journal.pone.0154566.g009 (CN) 6 electrolytes after 1500 cycles were 30%, 36%, and 44%, respectively. The addition of 0.4-M K 3 Fe(CN) 6 to the KOH electrolyte resulted in a small improvement in the electrode stability. The high loss of specific capacitance in this system could be attributed to substantial electrode degradation resulting from the high current passed through it during the cyclability test and volume loss of the active material [38,39]. Although the specific capacitance and energy density of this system were extremely high, its stability must be improved.

Conclusions
We have fabricated MnO 2 -NiO composite electrodes with a network-like structure using three electrodeposition modes: CP, CA, and CV. The composite CY7 electrode exhibited high homogeneity, a highly amorphous nature, and good capacitive behaviors compared to the CA and CP electrodes. The specific capacitances based on the mass of deposited MnO 2 -NiO in the CP, CA, and CY7 electrode were 435 Fg −1 , 458 Fg −1 , and 500 Fg −1 , respectively, and thus, CV electrodeposition was identified as the most suitable mode for MnO 2 -NiO composite electrode electrodeposition. The number of deposition cycles was also found to control the nucleation process, morphology, deposit thickness, and electrochemical performance of the resulting electrodes. Indeed, the capacitance decreased as the number of deposition cycles increased because of the weak wettability properties at the electrode/electrolyte interfaces and insufficient intercalation/deintercalation activity into or from the electrode. The best-performing electrode was found to be the CY4 electrode based on a comparison performed in Na 2 SO 4 electrolyte. The optimum electrochemical performance of the CY4 electrode was achieved in mixed KOH/K 3 Fe (CN) 6 electrolyte, resulting in a specific capacitance of 3509 Fg −1 and energy and power densities of 1322 Wh kg −1 and 110.5 kW kg −1 , respectively, at a current density of 20 Ag −1 . The simple electrodeposition of MnO 2 -NiO composite electrodes is cost efficient, versatile, and scalable for the synthesis of composite electrodes for supercapacitor applications.