The relentless increase in energy demand and parallel steady reduction of fossil fuels drive the scientific and industrial need to develop renewable energy along with faster and compatible energy storage devices (Li et al., 2015, 2017; Zhang et al., 2019). Supercapacitors represent essential components that can bridge between stereotype capacitors and rechargeable batteries in terms of power and energy densities with a long cycle life (Wang et al., 2018; Cao et al., 2019; Chen et al., 2019). In recent times, the synthesis of materials having semiconducting, metalloid, and energy storage properties has emerged at the forefront of research activities (Augustyn et al., 2014; López-Ortega et al., 2015; Sivula and Van De Krol, 2016). The transition metal phosphides are a fascinating class of materials because of their wide range of properties depending on their size and shape (Barry et al., 2008; Shao et al., 2017; Aziz and Islam, 2018). Metal phosphides have been tested and found to be active catalysts with their potential applications in electro-catalysis and photo-catalysis. In the last few years, several phosphides of nickel, cobalt, iron, molybdenum, tungsten, copper, and many more with different morphologies and crystallinity have been synthesized using different methods including phosphorization, electrodeposition, and ball milling (Brock et al., 2004; Henkes et al., 2007; Barry et al., 2008; Brock and Senevirathne, 2008; Wang et al., 2010). Metal phosphides with controlled nanostructures and elemental compositions have been designed and reported to provide significant enhancement in electrochemical activity (Wang et al., 2015a; Chen et al., 2016; Seo et al., 2016; Wu H. et al., 2016). Also, metal phosphides have metalloid properties and exhibit high specific capacitances (C_sp) and high volumetric and gravimetric capacities with lower intercalation potentials compared to commercial carbonaceous materials. Thus, they have attracted extensive interest to design novel electrode materials for supercapacitors rather than lithium-ion batteries (Li et al., 2017). Although transition metal hydroxides, metal oxides, and polymers (conducting) are commonly used electrode materials for lithium-ion batteries and pseudo-capacitors (Wang et al., 2015a, c; Zhu et al., 2015; Wu J. et al., 2016), however, these materials are kinetically slow for the fast electron transport and that invariably required for high-power density with moderate energy density (Li et al., 2017).

Copper phosphide (Cu₃P) is an air-stable, low-cost, environment-friendly material that exhibits high volumetric as well as comparable gravimetric capacity to that of graphite (Pfeiffer et al., 2004). Cu₃P have also been explored widely in the field of photo/electro-catalytic water splitting (Sun et al., 2015; Wei et al., 2016; Han et al., 2017; Hao et al., 2016). Recently, this material has been introduced in the energy storage sector to bridge the gap between electric double-layer capacitor (EDLC) and batteries (Chen et al., 2017; Zhang et al., 2019). It has been proved that the electrochemical activity of a material is strongly affected by its size, structure, and chemical composition. Cu₃P with different morphology and crystallinity have been synthesized and investigated for energy storage devices showing an improvement in electrochemical activities (Liu et al., 2012, 2016; Fan et al., 2016).

Carbonaceous matrices [e.g., graphene, carbon natotube (CNT), graphite, etc.] are simultaneously proved to be a promising choice of electrode material due to its high conductivity, large surface area, flexibility, and transparency which drags its attention toward large-scale applications in energy storage devices. Among various carbonaceous materials, three-dimensional graphene (3DG) is the promising choice as a backbone for electroactive materials for supercapacitor due to the random orientation of graphene flakes in three dimensions leading to the formation of micro/mesoporous structures (Mukherjee and Austeria, 2016; Singh et al., 2017; Xue et al., 2018). The 3DG provides high conductive channels for ion and electron transfer in charging–discharging cycles leading to a prominent improvement in overall capacitive performance of electroactive materials. Also, 3DG effectively minimizes the unwanted agglomeration of metal nanoparticles, thereby enhancing the chemical stability of the electroactive materials. In compliance to develop a novel approach for the synthesis of new hybrid materials (Kumar et al., 2017; Itzhak et al., 2018) herein, we report a facile, industrially scalable and easy method to produce a high yield of hexagonal Cu₃P by solid-state reaction at low temperature followed by the formation of nano-hybrid with 3DG matrix on graphite substrate. This nano-hybrid has demonstrated significantly high electrochemical performance at an optimized Cu₃P:3DG ratio with a tremendous C_sp of 1,095.85 F/g at a scan rate of 10 mV/s. The asymmetric device of Cu₃P@3DG electrode (positive) and activated carbon (AC) (negative electrodes) on graphite substrate reveals worthy performance in power densities, cycling stability, and excellent capacitive retention.

In this work, hexagonal platelets were directly grown on copper foil by thermal annealing using a CVD system. A known amount of red phosphorus was kept in a quartz boat and covered with copper foil as shown in Supplementary Figure S1. Red phosphorus and copper foil were then heated at 450°C for 5 min under argon flow. A dark shiny gray color layer that formed on the red phosphorus-exposed side of the copper foil indicated the successful growth of Cu₃P. To investigate the effect of temperature on the morphology of Cu₃P on copper foil, the same experiment was repeated at 350, 450, 600, and 800°C. When the temperature was increased to 600°C, the material started losing its hexagonal morphology by condensing hexagonal platelets together. Upon further enhancement of the temperature up to 800°C, the entire hexagonal morphology was lost. On the other hand, lowering the temperature to 350°C, the hexagonal platelets did not nucleate, as shown in Supplementary Figure S2. These results confirmed that a specific temperature (450°C) was required for the growth of hexagonal platelets. HRSEM analysis of the synthesized Cu₃P revealed stacked platelets of hexagonal morphology with a lateral dimension in the range of micrometers and a thickness of a few nanometers, as shown in Figure 1A. HRSEM images and the corresponding elemental mapping patterns of Cu and P had demonstrated that the Cu and P elements were uniformly distributed on the hexagonal platelets (Supplementary Figures S3c,d in ESI). Transmission electron microscopy (TEM) analysis was performed for in-depth characterization. A low-magnification image, shown in Figure 1B, illustrated the nature of the hexagonal platelets. It was found that they consist of domains, although TEM sample preparation might influence the agglomeration (Supplementary Figure S4). The electron diffraction pattern taken from the agglomerate portion (Supplementary Figure S4a) was successfully indexed in terms of the hexagonal structure of the Cu₃P. It should be noted that the rings on the polycrystalline electron diffraction pattern indicated a large misorientation between the domains in the agglomerate. Using HRTEM, we studied small nano-sized domains constituting the agglomerate and did selected area electron diffraction (SAED), shown in Figures 1C–D. The SAED pattern was taken along [0001] orientation of one domain, and the corresponding HRTEM image is shown in Supplementary Figure S4b. In Supplementary Figures S4b,c, HRTEM image and SAED patterns were taken along [101 ^–0] orientation of the domain fitting the atomic structure of Cu₃P, as reported by Olofsson (1972).

The phase content of the synthesized material was studied using X-ray diffraction (XRD) (Figure 1E). The XRD pattern exhibited strong, sharp peaks which were indexed in terms of hexagonal structure (P63cm) of the Cu₃P (JCPDS card no. 71-2261) (Hou et al., 2016). Rietveld refinement resulted in reliability factors: R_p = 2.3%, R_wp = 3.27%, R_Bragg = 1.05%, and R_exp = 1.73%. These low-reliability values excluded the possibility of the presence of impurities. It should be noted that atomic positions, occupancies, and thermal displacement parameters of the Cu₃P structure were not refined here, and the background was treated as linear. Lattice parameters were refined as a = 0.69673(1) nm and c = 0.71484(8) nm. The dependence of stoichiometry, temperature, and lattice parameters of the Cu₃P structure was discussed in the earlier report (Olofsson, 1972). Since the refined lattice parameters (i.e., a = 0.697, c = 0.7145 nm) were fitted well with reported parameters (Olofsson, 1972), thus the synthesized material was stoichiometric (or closely stoichiometric) as Cu₃P. To the best of our understanding, Rietveld refinement result of the as-synthesized sample was known to be the significant evidence for the formation of stoichiometric Cu₃P rather than the energy-dispersive X-ray spectroscopy (EDS) shown in Supplementary Figure S3. The presence of elements could be found in the EDS analysis. However, it was hard to estimate the stoichiometry via EDS since a large difference in atomic numbers of the constituents did not allow accurate standard in the EDS analysis. Figure 1F showed the formation of a hybrid porous structure made of hexagonal Cu₃P platelets along with 3DG. Careful observation indicated that the hexagonal Cu₃P platelets (shown by yellow arrows) were randomly scattered on the porous foamy 3DG architecture (shown by white arrows). A close-up view of these hybrid structures, shown in Figure 1G, revealed a single flake of Cu₃P trapped inside the 3DG matrix where a detailed EDS elemental mapping was performed to identify the distribution of Cu, P, and C elements, respectively, shown in Supplementary Figure S5.

The X-ray photoelectron spectroscopy (XPS) analysis of Cu₃P elucidated the oxidation states and chemical interaction between Cu and red phosphorus (Supplementary Figure S6). The Cu, P, O, and C peaks appeared in the survey scan (shown in Supplementary Figure S6a). The presence of O1s peak indicated the superficial oxidation of Cu₃P, while the carbon peak was due to the background. The high-resolution XPS spectrum of Cu 2p (Supplementary Figure S6b) exhibited two peaks at 932.4 and 934.8 eV, which were attributed to phosphorized copper (Cu–P) and oxidized copper (Cu–O), respectively (Lin et al., 2016; Jin et al., 2017). Supplementary Figure S6c showed double peaks in the phosphorus region. The peak at 133.6 eV was due to the oxidized phosphorus in the form of phosphate (Li et al., 2016), and 129.7 eV was the binding energy of P in Cu₃P, which was lower than that of the binding energy of red phosphorus (130.2 eV), suggesting the conversion of copper to Cu₃P (Pfeiffer et al., 2005). The presence of Cu–O and POx was probably due to oxide layer formation on the surface of Cu₃P in ambient conditions.

To evaluate the as-grown nanomaterial as the promising electrode material, cyclic voltammetry (CV), galvanostatic charge–discharge (GCD), and electrochemical impendance spectroscopy (EIS) measurements were made to investigate the supercapacitive performance. The electrochemical window for CV and GCD measurements was kept in −0.5 to +0.5 V shown in Figures 2A,B. Figure 2A showed CV of the Cu₃P nanoplatelets under a three-electrode system in 1 M Na₂SO₄ electrolyte at a different scan rate from 10 to 200 mV/s. The broad shape of the asymmetric CV curves observed for Cu₃P nanoplatelets in the whole potential window indicated the presence of faradic behavior leading to pseudocapacitance along with electric double-layer capacitance (EDLC) existing simultaneously.

The appearance of two humps in between −0.5 and +0.5 V corresponding to cathodic and anodic maxima indicated that the Cu₃P platelets possess pseudocapacitive properties. These redox reactions attributed to the transition between metallic Cu(0) and Cu(I) species at an applied potential approximately −0.3 V vs. Ag/AgCl. While at the higher potential of 0.05 V vs. Ag/AgCl, the oxidation of Cu₃P platelets to Cu(II) species was observed due to aqueous interaction (Figure 2A). Further, the broad anodic peak nearly at 0.3 V could be attributed to the formation of a hydrous oxide of copper which originated from the hydrated cupric oxide (Jayalakshmi and Balasubramanian, 2008), and the corresponding chemical reaction was shown in equation (3). Here, the potential window had been fixed to 0.5 V to avoid the formation of Cu(III) species. So far, the identification of intermediate species was hard to find due to the chemical transformation of Cu₃P electrodes under electrochemical analysis, and the reaction mechanism was entirely unknown which required further detailed investigation. However, based on the available relevant literature, a possible reaction mechanism had been proposed in the following equations (Poli et al., 2016; Chen et al., 2017).

This mechanism could be attributed to redox reactions on the surface of Cu₃P platelets. Pfeiffer et al. (2004) had explained that all the elements of the first transition metal series have partially filled 3d shells, except copper and zinc. Nonetheless, only the copper has a complete 3d shell and a single 4 s electron. Hence, copper was the only element in the series to have M⁺¹ state with filled D-orbital in Cu₃P platelets. The enhanced electrochemical performance might be attributed to the fact that the hexagonal Cu₃P platelets were composed of many small particles, which had significantly increased the specific surface area and provided more reaction sites on the surface evident from the SEM image shown in Figures 1F,G.

To further improve the Cu₃P electrode performance, we had designed a hybrid electrode with 3DG. The homogeneous distribution of Cu and P in hexagonal Cu₃P flake and carbon in the foamy graphitic matrix had confirmed the formation of Cu₃P@3DG hybrid network (shown in Supplementary Figure S5). For the development of effective electrode material, it was highly required to estimate the electrical conductivity of the proposed electrodes. The electrical conductivities of the Cu₃P@3DG and 3DG electrodes via 4PP configuration with the electrode spacing of 2 mm were calculated to be 1,076 S m^–1 and 978 S m^–1, respectively. Here, the Cu₃P@3DG electrode exhibited better conductivity as compared to the bare 3DG electrode on graphite substrate. This could be due to the metalloid property of Cu₃P nanoparticles, shown in Supplementary Figure S7. To boost the kinetics of Cu₃P, Cu₃P@3DG had been elaborately designed, as schematically shown in Figure 1A. A systematic loading optimization of Cu₃P on a 3DG matrix had been performed for extracting the best capacitive performance of Cu₃P@3DG electrodes, as shown in Figures 2C,D. Here, the area under the CV curve was gradually increasing without any impressive redox peaks under the loading concentration of Cu₃P that varied from 0.3 to 0.91 mg/ml. Optimizing the weight% of Cu₃P in 3DG electrode, the redox probability of Cu₃P on the 3DG matrix had been controlled until 0.91 mg/ml loading. However, increasing the loading concentration to 1.2 mg/ml and above, the redox peaks appeared more prominent at the same potential as bare Cu₃P with increased current density due to the presence of the 3DG matrix. However, after a further increase in the concentration of Cu₃P until 1.5 mg/ml, the prominent redox peaks appeared. This loading concentration of Cu₃P on 3DG-graphite (0.91 mg/ml loading in 1 cm² × 1 cm²) showed the best performance in galvanostatic charge–discharge curves at 5.13A/g current density (Figure 2D). For device fabrication as well as to evaluate the stability of the supercapacitor electrodes, the optimum loading concentration of Cu₃P had been fixed to 0.91 mg/ml. The electrochemical CV curves and GCD performance (with different scan rates) of 3DG and bare Cu₃P on graphite substrate are shown in Supplementary Figure S8. However, it had been observed that the C_sp value of Cu₃P on glassy carbon electrode (234.8 F/g) was slightly higher than that on graphite electrode (224 F/g) at a fixed current density of 1.28 A/g, which might be due to the interfacial interaction of electroactive material with the electrode substrate. The electrochemical analysis using this hybrid structure was carried out to evaluate its applicability toward energy storage devices. The C_sp of Cu₃P@3DG electrode was calculated to be 1,095.85, 901.50, 828.48, 778.70, 740.95, 706.66, 675.37, 590.95, 621.63, 597.45, 574.29, 495.97, and 421.65 F/g at the scan rate of 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, 150, and 200 mV/s, respectively from Figure 2E. Alternatively, the C_sp extracted from the GCD measurements (Figure 2F) of the Cu₃P@3DG electrode at various current density were calculated to be 849.81, 771.26, 734.35, 712.35, 692.39, 676.9, 659.98, and 644.75 F/g at 1.28, 2.56, 3.85, 5.13, 6.41, 7.69, 8.97, and 10.268 A/g, respectively (Supplementary Figure S9). The nearly isosceles triangular shape of the GCD (Figure 2F) implied excellent electron conductivity with tiny voltage drops at different current densities of Cu₃P@3DG electrode. The energy and power density extracted from the charge and discharge analysis were about 118.1, 107.1, 102.0, 99.0, 96.2, 94.1, 91.7, and 89.5 Wh/kg and 641.0, 1,282.0, 1,923.0, 2,564.1, 3,205.2, 3,846.2, 4,487.2, and 7,132.2 W/kg at 1.28, 2.56, 3.85, 5.13, 6.41, 7.69, 8.97, and 10.268 A/g current density, respectively.

The comparative CV curves of the electrodes made of three different electroactive nanomaterials: Cu₃P@3DG, 3DG, and bare Cu₃P were carried at a scan rate of 20 mV/s (Figure 3A). The C_sp of Cu₃P@3DG, 3DG, and Cu₃P electrodes were calculated to 901.5, 332.7, and 217.7 F/g (from CV curves), respectively, at a scan rate of 20 mV/s. Typically, the electrochemical properties like C_sp of Cu₃P@3DG were calculated to be nearly 2.5 times and four times enhanced as compared to bare 3DG and Cu₃P-based electroactive materials, respectively. The results implied that Cu₃P@3DG electrode referred to an excellent EDLC. The GCD measurements of the Cu₃P@3DG, 3DG, and bare Cu₃P on graphide electrodes at fixed current densities were performed with the potential window of 1.0 V (Figure 3B). The Cu₃P@3DG exhibited a high C_sp of 850 F/g (from GCD analysis) as compared to 3DG and bare Cu₃P of 336 F/g and 224 F/g, respectively, at a current density of 1.28 A/g. Even at a very high current density of 10.27 A/g, the C_sp reached up to 645 F/g which was about 76% of the capacitance retention compared to the current density of 1.28 A/g that invariably referred to the excellent stability of our electroactive material hetero-structure (shown in Figure 2F). The primary objective to incorporate 3DG along with Cu₃P platelets was to provide high conductive channels for ion and electron transfer in the charging–discharging process leading to the improvement of the overall performance of hybrid electrodes (An et al., 2014). This hybrid structure could effectively minimize the unwanted agglomeration of Cu₃P particles and simultaneously restricted the possibility of restacking of 3DG porous structure and thereby retaining the large surface area (Supplementary Figure S10), higher C_sp, better rate capability, and overall device stability. However, the bare 3DG (made of three-dimensionally oriented rGO with very low oxygen content shown in Supplementary Figure S11) is known to be acting as promising anode materials due to high electronegativity rather than effective cathode material where graphene is difficult to get oxidized under the operating potential window (Hossain et al., 2012). CV analysis of bare glassy carbon and Cu₃P decorated glassy carbon shown in Supplementary Figure S12.

The cycle stability had been evaluated for Cu₃P@3DG, 3DG, and hexagonal Cu₃P platelet-based electroactive materials on graphite substrate for its applicability in real phase energy storage devices. Figure 3C showed the variation of the C_sp retention (%) as a function of the cycle number of the galvanostatic charge–discharge curves of the Cu₃P@3DG, 3DG, and hexagonal Cu₃P platelets at a fixed current density of 1.28 A/g. Surprisingly, after 3,000 cycles, the capacitive retention of Cu₃P@3DG was found to be 95%, whereas that of 3DG was 83%. However, the stability degraded to 75% for bare Cu₃P platelets. This gave us a clear understanding that the 3DG matrix was protecting Cu₃P platelets from chemical degradation as well as agglomeration due to its high electrical and thermal conductivities. These excellent properties of 3DG in our nanohybrid materials had been further evaluated by characteristic electrochemical impedance study. Figures 3D–F showed the Nyquist plot of the Cu₃P@3DG, bare 3DG, and Cu₃P electrodes, respectively. In the spectrum, the equivalent series resistances of Cu₃P@3DG, 3DG, and bare Cu₃P on graphite electrodes were calculated to be 1.6, 8.9, and 18.5 Ω, respectively, which showed favorable conductivity (evident from the 4PP I-V characteristic, shown in Supplementary Figure S7) and very low internal resistance of Cu₃P@3DG hybrid for high specific power density.

The asymmetric supercapacitor (ASC) device was fabricated to investigate the potential application of Cu₃P@3DG nanohybrid as a positive electrode and AC as a negative electrode on a graphite substrate. The electrochemical performance of the ASC device was analyzed via CV at different scan rates, and GCD measurements were done at various current densities, as shown in Figures 4A,B. The potential window of the ASC Cu₃P@3DG and AC electrodes was fixed to be 0–1 V. Figure 4A displayed the CV curves of the ASC device at different scan rates with a constant potential window, which showed rectangular like CV curves without redox peaks. It indicated the excellent reversibility and kinetics for electrochemical reactions over the ASC electrode. The C_sp of the device was calculated to be 108.78, 61.65, 47.30, 39.86, and 35.09 F/g (217.56, 123.32, 94.61, 79.74, and 70.20 mF/cm²) at the scan rate of 10, 50, 100, 150, and 200 mV/s respectively. Alternatively, the C_sp extracted from the GCD measurements of the device, as shown in Figure 4B, at various current densities were calculated to be 59.72, 51.82, 39.53, and 37.53 F/g (107.8, 94.33, 71.96, and 68.32 mF/cm²) at 0.5, 0.6, 0.7, and 0.8 mA/cm² current densities, respectively. The energy and power densities were calculated to be 8.23, 7.2, 5.5, and 5.2 Wh/kg and 274.7, 329.7, 384.6, and 439.6 W/kg at 0.5, 0.6, 0.7, and 0.8 mA/cm² current densities, respectively. To further elucidate the origin of high electrochemical performance, the electrochemical impedance spectrum (EIS) was carried out to examine the Cu₃P@3DG//AC supercapacitor device. Figure 4C showed the Nyquist plot, the combined resistance Rs (the intrinsic, contact, and electrolyte resistance) of 4.24 Ω, the Cdl (the electric double layer capacitance) of 70.4 μF, and the Rct (the resistance of Faradic reaction) of 714 mΩ. The inclined line in the low-frequency region represented the Warburg impedance (W) of 19.2 mMho, and constant phase element (C_F) 81.1 mMho. Simultaneously, the apparent straight-line nature of the plot in a low-frequency region described the ideal capacitive behavior of this prototype ASC device. The corresponding equivalent circuit was drawn based on the complex non-linear least-squares fitting method, shown as an inset of Figure 4C. Figure 4D displayed the Bode phase angle plot of Cu₃P@3DG//AC. The absence of a horizontal segment in the Z-frequency plot at the low-frequency range revealed good charge-transfer behavior. Simultaneously, the high value of the phase angle curve manifested the great capacitive behavior of Cu₃P@3DG//AC at the low-frequency range. Moreover, the long-term cycling stability performance was examined by GCD analysis at a current density of 0.7 mA/cm². This supercapacitor device has demonstrated 82.5% capacitance retention after 5,500 cycles exhibiting outstanding cycling stability with 96% coulombic efficiency, as shown in Figure 4E. Figure 4F displayed the Ragone plot of ASC (Cu₃P@3DG//AC) device. All these observations indicated that the Cu₃P@3DG exhibited good electrochemical performance that could promote it to be an excellent material for its applicability in the next-generation energy storage device. In all the listed combination of metal phosphides, a direct comparison was not possible because of various factors like two-electrode experiment was performed in both solid and liquid state and the concentration and type of elecrolytes were different in all experimental conditions which had a prominent effect on the overall performance of the materials.

Table 1 depicts the transition metal phosphide-based active electrodes that so far had been tested as promising materials for supercapacitor devices. The synthesis of the metal phosphides with optimized nanostructures and composition was a real challenge. Although the reported specific capacitances of the active materials like Ni-P, Ni₂P, CoP, Co₂P, and Cu₃P with different morphologies were quite high, but the working potential and capacitance retention were significantly low. However, it was found to be quite challenging to expand the working potential of the electrodes for practical applications. Here, we offered a strategy to synthesize hexagonal Cu₃P platelets with 3DG as promising electrode materials for supercapacitors via controlling the directionality of the growth as well as engineering the facile interface between electrodes to electroactive materials which worked well in negative as well as positive working potentials.

In summary, the facile synthesis procedure of hexagonal Cu3P platelets and the fabrication of a heterostructure Cu₃P with interwoven 3DG on graphite electrode (1 cm² × 1 cm²) shows a simple way to design a supercapacitor device for practical applications. The Cu₃P@3DG nanohybrid facilitates electron transfer and promotes kinetics with long cycling stability via lowering the internal resistance and providing electrochemical stability of Cu₃P platelets. The high conductive channels for ions and electron transfer in the charging–discharging process leads to an enhancement of the overall device performance. The specific capacitance of the Cu₃P@3DG is 849.81 F/g at a current density of 1.28 A/g and also exhibits superior cycling performance, with 95% retention of capacitance after 3,000 cycles at a current density of 1.28 A/g. In contrast, the solid-state asymmetric device shows the C_sp of 109 F/g (from the CV) and 60 F/g (from GCD) with a capacitive retention of 82.5% and coulombic efficiency of 96% after 5,500 cycles, respectively. Due to the higher capacitance, lower cost, and excellent cycle stability of the Cu₃P@3DG, it is anticipated that the material has potential for next-generation energy storage applications.

All datasets generated for this study are included in the article/Supplementary Material.

SBK designed the work. SSK and SA performed the experiments and characterizations. SSK and SR prepared the electrodes assembly of the samples and analyzed the electrochemical data. KG and GN participated in interpreting and analyzing all data and finalized the manuscript. GY and LM helped to characterize HRTEM and TEM analysis. All authors read and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.