Article | Open | Published:

# Flexible Fe3O4@Carbon Nanofibers Hierarchically Assembled with MnO2 Particles for High-Performance Supercapacitor Electrodes

## Abstract

Increasing use of wearable electronic devices have resulted in enhanced demand for highly flexible supercapacitor electrodes with superior electrochemical performance. In this study, flexible composite membranes with electrosprayed MnO2 particles uniformly anchored on Fe3O4 doped electrospun carbon nanofibers (Fe3O4@CNFMn) have been prepared as flexible electrodes for high-performance supercapacitors. The interconnected porous beaded structure ensures free movement of electrolyte within the composite membranes, therefore, the developed supercapacitor electrodes not only offer high specific capacitance of ~306 F/g, but also exhibit good capacitance retention of ~85% after 2000 cycles, which certify that the synthesized electrodes offer high and stable electrochemical performance. Additionally, the supercapacitors fabricated from our developed electrodes well maintain their performance under flexural stress and exhibit a very minute change in specific capacitance even up to 180° bending angle. The developed electrode fabrication strategy integrating electrospinning and electrospray techniques paves new insights into the development of potential functional nanofibrous materials for light weight and flexible wearable supercapacitors.

## Introduction

Owing to the rapidly developing market for high performance portable electronic devices, design and fabrication of flexible energy management devices has been very critical1,2,3,4. In this regard, scientists are paying special attention to the supercapacitors which owe the enough potential to offer high energy density, superior power density, and excellent capacitance retention compared to dielectric capacitors and rechargeable batteries5,6. Various carbonaceous materials, such as graphene and carbon nanotubes etc., have been employed as active electrode materials for developing flexible supercapacitors owing to their high specific surface area and outstanding electrochemical and mechanical properties7. However, their high cost and complex fabrication process urge scientists to dig out alternative economical materials with superior mechanical characteristics and excellent flexible structure8. On the other hand, carbon nanofibers (CNF) fabricated via electrospinning (an easy, economic, and scalable production process9,10) offer utmost surface area and tailorable intra- and inter-fiber pore structure, therefore, are believed to be potentially ideal materials for developing high performance flexible supercapacitor electrodes11. However, pristine CNF tend to be relatively fragile in nature, therefore, certain additives for proper transmittance and scattering of external forces are needed to enhance their firmness against bending forces12,13,14.

In this regard, transition metal oxides (TMOs, compounds of oxygen bound to the transition metals), could be useful additives when employed with electrospun CNF to form nanoclusters. The uniform distribution of these nanoclusters within fiber matrix can transmit and scatter the external stresses, reduce the magnitude of the applied stress per unit area, thus, act as a barrier to crack generation and propagation15,16. Besides that, pore deformation and slips of graphitic carbon layers also disperse and absorb external forces, hence, support the structural flexibility of CNF17. Additionally, they could also be used to smartly tailor the electrochemical characteristics for developing conductive flexible electrospun CNF for portable supercapacitor applications18,19.

Recently, numerous TMOs (NiO20, CoOx21, MnO2 22, V2O5 23 etc.) offering relatively low cost, abundant availability and considerable electrochemical behaviors have been reported24,25,26. Among these TMOs, Fe3O4 is gaining special attention owing to its low cost, environment friendly nature, and readily availability27. However, excessive use of Fe3O4 may lead to the high volumetric changes in supercapacitors when employed as pseudo-capacitive material28. Additionally, its low ionic conductivity also hinders its extensive use in supercapacitors as a conducting additive29. Furthermore, Fe3O4 has certain limitations to be used as an electrochemical additive, however, optimized concentration of Fe3O4 may yet be very useful to tailor mechanical characteristics and flexibility of CNF without compromising their electrochemical characteristics30,31. Whereas, MnO2 besides being economical and environment friendly material, also offers excellent specific capacitance. Moreover, abundant availability and easier applicability via multiple fabrication processes (electrospinning, thermal decomposition, and chemical vapor deposition) make it a potential additive material for supercapacitor applications32,33. Therefore, a combination of Fe3O4 and MnO2 34 can develop a prospective material for fabricating flexible CNF with high electrochemical performance for supercapacitor applications35.

In the current work, we report a highly flexible Fe3O4 doped CNF/MnO2 (Fe3O4@CNFMn) supercapacitor electrode with superior electrochemical performance. Fabrication process initiated with electrospinning of a homogenous solution comprising of polyacrylonitrile (PAN) and Fe(acac)3 dissolved in dimethyl formaldehyde (DMF), followed by preoxidation and carbonization for their successful conversion into Fe3O4@CNF. The presence of Fe3O4 in CNF turned them into a highly flexible membrane. Later on, MnO2 deposition via electrospray not only enhanced the electrolyte uptake of the resultant Fe3O4@CNFMn membrane, but also boosted the overall electrochemical performance.

## Results

### Morphology and structure analysis

Figure 1 demonstrates the schematic illustration of the fabrication pathway of Fe3O4@CNFMn composite membrane. At first, an identical solution comprising of PAN and iron acetylacetonate (Fe(acac)3) dissolved in DMF was electrospun. Resultant fibers tagged as Fe@PAN (Figure S1) were vacuum dried and stabilized by curing at 280 °C. Later on, stabilized Fe@PAN nanofibers were successfully carbonized (800 °C) under inert environment and corresponding samples were tagged as Fe3O4@CNF. Earlier, in our previous work, we figured out that pure CNF are relatively fragile in nature and have relatively low ionic conductivity and affinity towards electrolyte11, thus, need some metallic crystallites which may be helpful to scatter the applied forces to prevent crack development, and conductive additives to tailor their ionic conductivity and electrolyte uptake capability to obtain the desired electrochemical performance36. Therefore, first objective of the current work to acquire desired flexibility in CNF was achieved by inducing the regulated concentration of Fe3O4 particles in the fiber matrix which help to scatter bending stress uniformly along fiber axis, acted as a barrier to applied stress, hence, inhibited the crack propagation in the resultant fibers. Consequently, the resultant carbonized samples showed extremely high flexibility (Movie 1) when compared to pure CNF (Movie 2). Furthermore, in order to achieve second objective of high electrochemical performance, the surface of the prepared Fe3O4@CNF composite membrane was first modified with 0.1 M H2SO4 followed by electrospray of KMnO4 37 (i.e. potential energy storage and low cost material with abundant availability and environment friendly nature), then washed with water to remove extra KMnO4 from the surface of the membrane and dried at 50 °C for 2 h. The electrosprayed KMnO4 was then converted into MnO2 particles via redox reaction38, which induced good ionic conductivity39.

Figure 2a,b presents the surface morphology of Fe3O4@CNF and Fe3O4@CNFMn. It could be observed that both membranes offered randomly organized three dimensional fibrous structure and inherited cylindrical shaped fiber structure similar to the precursor membranes. Owing to very low concentration of Fe3O4, Fe3O4 particles were uniformly deposited inside as well as on the exterior wall of the resultant fibers (Figure S2), and did not significantly affect the fiber diameter of the subsequent Fe3O4@CNF membranes. The average fiber diameter of Fe3O4@CNF membrane ranged from 390 ± 20 nm, and the resultant membranes offered relatively low ionic conductivity, therefore, in order to achieve the desired ionic conductivity, subsequent membranes were electrosprayed with a relatively high ionic conducting material (i.e. MnO2). Random deposition of KMnO4 particles after redox reaction formed bead like MnO2 crystals on the surface of Fe3O4@CNF (Fig. 2b), enhanced the surface roughness and created open inter-particle pores on the surface of Fe3O4@CNF. Additionally, hydrophilic nature of MnO2  40 may lead to higher electrolyte uptake, thus, would improve ionic conductivity and specific capacitance of resultant Fe3O4@CNFMn composite membranes. In order to investigate the electrolyte uptake behavior, Fe3O4@CNF and Fe3O4@CNFMn were subjected to WCA examination for further analysis. Electrolyte uptake capability of the two membranes was analyzed by examining the static WCA. Fe3O4@CNF did not allow water droplets to spread much on surface and displayed a highly hydrophobic nature with the WCA of 130°, and the WCA magnificently dropped to 40° after electrospray with MnO2 for Fe3O4@CNFMn (Figure S3), indicating higher potential for electrolyte uptake compared to Fe3O4@CNF.

Moreover, the presence of Fe3O4 and MnO2 is also confirmed by TEM analysis (Fig. 2c,d). TEM image shows that nanoclusters were distributed throughout the surface of the fiber, i.e. edges as well as the body of the fiber (Fig. 2c). However, owing to deposition of a thick layer of MnO2 on the synthesized Fe3O4@CNFMn composite membranes, Fe3O4 sights have been overlapped with MnO2 crystallites, thus, are not very obvious from TEM image. Therefore, HR-TEM examination was employed to validate the presence of Fe3O4 particles in the resultant membrane (Fig. 2d), which confirmed the presence of Fe3O4 sights in the final product by showing an inter-plane distance of 2.6 Å in between the well-ordered fringes41. Furthermore, presence of Fe3O4 particles was also confirmed by cross-sectional SEM image (Figure S2), and close observation reveals that Fe3O4 is well dispersed in the fiber matrix with low agglomeration, hence, played lead role in developing flexible CNF.

Nitrogen adsorption-desorption isotherms were employed to investigate the porosity and surface area of the resultant Fe3O4@CNF and Fe3O4@CNFMn (Fig. 3a). Presence of micro- as well as mesopores for two membranes can be seen from the weak hysteresis of the typical isotherms at high relative pressure. Besides that, pore size distribution obtained by using nonlocal density functional theory (2-NLDFT) as shown in Fig. 3b, also validates the existence of micro- (Figure S4) and mesopores. Moreover, moderate BET specific surface area (~162 and ~148 m2/g) and considerable pore volume (~0.268 and ~0.254 cm3/g) for two membranes i.e. Fe3O4@CNF and Fe3O4@CNFMn, was observed. MnO2 deposition probably blocked some of the surface pores resulting in a minute decrease in surface area and pore volume.

Raman spectroscopy was used for analyzing the nanostructures and degree of graphitization of the subsequent composite membranes in Fig. 3c. Two characteristic peaks at 1344 and 1580 cm−1 corresponding to D and G bands, respectively, show vibrations with sp3 bonds in the crystal lattice defects and stretching of carbon atoms bonded with sp2 bonds. Sp3 bond vibrations lead to disordered carbonaceous matrix, whereas stretching of carbon atoms bonded with sp2 bonds could be credited for ordered graphitized structure. Furthermore, the standard XRD patterns of Fe3O4, MnO2 and carbon well matched with reported diffraction peaks19,21, which indicate the successful development of Fe3O4 and MnO2 in the resultant composite CNF membrane (Fig. 3d). Moreover, the corresponding peaks at 002 and 100 represent carbon, and peaks at 001 and 311 show the presence of Fe3O4, whereas peak at 201 represent to MnO2. Chemical composition and structural characteristics of synthesized composite membrane are also validated by FTIR spectra (Figure S5). The characteristic peaks at 468, 518 and 1405 cm−1 correspond to Mn-O stretching, Fe-O stretching and C-H stretching, respectively, whereas peaks at 1163 and 2965 cm−1 represent the stretching mode of C-O and N-H, respectively.

Based on the systematic analysis of the nanostructures of CNF membranes, the possible mechanism behind enhanced flexibility of synthesized CNF is demonstrated in Fig. 4. Ordered distribution of the Fe3O4 and ordered graphitized carbon layers in the fiber matrix form a hierarchical composite nanostructure. Owing to the interconnected net like structure of CNF membranes, the external stress on the CNF membranes would lead to structural deformation of single fibers. Bending stresses remain condensed on the subjected area in the pure CNF, therefore, rapidly overcome their tolerance against external stress and result in formation and propagation of cracks leading to the complete fracture of membrane42. Whereas, incorporation of the Fe3O4 crystallites and MnO2 beads distribute the applied bending stresses and scatter them rapidly in fiber axis, hence, magnitude of the force per unit area is decreased43. Moreover, this deposition of Fe3O4 and MnO2 also absorb some of the applied forces and act as buffer to stop extension of tiny cracks. Therefore, composite CNF can withstand higher magnitude of bending forces and show high flexibility.

### Electrochemical performance

The electrochemical performance of the developed membranes was evaluated by using galvanostatic charge/discharge (GCD), cyclic voltammetry (CV), and electrochemical impedance spectroscopy (EIS). Since the sole purpose of using metal particles was to develop flexible carbon nanofibers, therefore, optimized quantity of metallic particles was incorporated in the current work to induce flexibility and avoid volumetric changes in the synthesized CNF. Thus, Fe3O4 presence had just negligible effect on the electrochemical performance of the synthesized membrane which is also evident from Figure S6.

Figure 5a presents the CV curves of the Fe3O4@CNFMn at various scan rates (10–30 mV/s) in the voltage range of 0–0.5 V. To further elucidate the electrochemical behavior of the Fe3O4@CNFMn, GCD test was carried out at various current densities 1–8 A/g (Fig. 5b and Figure S7). GCD curves in the potential range endorsed the high charge-discharge reversibility, however, specific capacitance for two membranes decreased with increment of current densities (Fig. 5c). Specific capacitance values obtained for Fe3O4@CNF were 120, 98 F/g at 1 and 2 A/g (Figure S8), and for Fe3O4@CNFMn were 306, 196, 143, 103 F/g at rising current densities of 1, 2, 4 and 8 A/g, respectively (Fig. 5c). It is worth noting that only a minute MnO2 loading (39 mg/g) on Fe3O4@CNF could result in the highest specific capacitance of ~306 F/g at the current density of 1 A/g. Furthermore, specific capacitance values for corresponding GCD curves (equation 1) for Fe3O4@CNFMn are not only relatively much higher than Fe3O4@CNF but also higher than many of recently reported works30,44. A comprehensible comparison of current work with previous reported works is illustrated in Table S1. The specific capacitance of these flexible Fe3O4@CNFMn can be calculated by using the following equation:

$${\boldsymbol{C}}=\frac{{\boldsymbol{I}}{\rm{\Delta }}{\boldsymbol{t}}}{{\boldsymbol{m}}{\rm{\Delta }}{\boldsymbol{V}}}$$
(1)

where, I, Δt, m and ΔV represent current, discharge time, mass of electrodes, and voltage, respectively.

Furthermore, electrochemical impedance spectroscopy (EIS) was used to comprehend the mechanism responsible for such an excellent specific capacitance performance of Fe3O4@CNFMn composite membrane, and is represented by the Nyquist plot. As shown in Fig. 5d, Nyquist plot could be differentiated into two regions, i.e. semicircle located in high-frequency region and Nyquist curve (Warburg impedance) in the low frequency region. Semicircle represents the electrolyte resistance (resistance of electrode and solution), therefore, smaller the diameter of semicircle in the EIS spectrum, the lower would be the charge-transfer resistance. Inset of Fig. 5d shows the magnified image of semicircle region. The straight line witnessed along the imaginary axis corresponds to the polarizable capacitance15. Moreover, no significant IR drop in GCD curves also validates the ideal capacitive behavior of the synthesized composite membrane which could be credited to the higher ionic conductivity and relatively improved electrolyte uptake as compared CNF45.

Figure 5e illustrates the variation in specific capacitance with respect to GCD cycle numbers, minute decrease in specific capacitance for the developed composite membranes was observed for first 300 cycles which could be attributed to the linear activation of Fe3O4 and MnO2, however, there was enough change observed in specific capacitance after 2000 cycles, but was not enough to alter the performance of supercapacitor (Fig. 5e, inset presents the corresponding GCD curves of 15 cycles). Additionally, prepared membranes showed reasonably high capacitance retention (i.e. ~85% of its initial value) after 2000 cycles (Fig. 5f, inset shows the digital image of flexible Fe3O4@CNFMn electrode), however, reason of 15% drop in specific capacitance could be credited to the MnO2 dissolution, oxygen evolution reaction (common drawback of MnO2) and structural deformation at nanoscale. In order to confirm structural deformation at nanoscale, samples were further subjected to FTIR and XPS examinations after cyclic performance. FTIR analysis of Fe3O4@CNFMn after cyclic performances showed that there was obvious decrease in the 468 cm−1 region which confirmed the MnO2 dissolution (Figure S5). Elemental analysis via XPS also validated the above findings. Existence of C, O, Fe and Mn in the Fe3O4@CNFMn samples (before and after cycles) could be endorsed from Fig. 6. High resolution XPS spectra of individual elements are illustrated in Figure S9. The binding energy separation of 11 eV between the peaks at 653 and 642 eV attributed to Mn 2p1/2 and Mn 2p3/2, respectively, and separation energy of 4.9 eV of Mn 3 s spin orbit doublet45 confirmed the intermediate state of Mn (Figure S9). It could be observed that Mn 2p1/2 and Mn 2p3/2 peaks declined after cyclic performance which validated the structural deformation of Mn at nanoscale, thus, may be credited for 15% loss of specific capacitance for after 2000 GCD cycles.

The fabricated supercapacitor device comprised of gel electrolyte in the center covered with Fe3O4@CNFMn composite membrane on each side followed by a gold sputtered polyethylene terephthalate (PET) protective layer (Fig. 7a). As the objective was to design flexible supercapacitors, therefore, feasibility of optimized samples (Fe3O4@CNFMn) for developing flexible supercapacitor devices was evaluated by investigating the specific capacitance of the fabricated symmetric supercapacitor device when subjected to bending force (Fig. 7b). No significant change in specific capacitance for optimized electrode even at bending angle of 180° was figured out (Fig. 7c), which confirmed that the synthesized composite nanofiber membranes have enough potential to fabricate flexible supercapacitors and could be used for wearable applications. Additionally, physical stability against flexural forces was also investigated using bending deformation cycles to determine the extent of flexibility of developed supercapacitors (Figure S10). It was figured out that even after 1000 bending cycles at 180°, there was no significant change in specific capacitance as well as physical structure of samples which ensured long and stable life of synthesized membrane.

Energy and power densities are two key characteristics which determine the energy storage and charging potential of supercapacitors, and are calculated by following equations.

$${\boldsymbol{E}}=\frac{1}{2}{\boldsymbol{C}}{{\boldsymbol{V}}}^{2}$$
(2)
$${\boldsymbol{P}}=\frac{{\boldsymbol{E}}}{{\rm{\Delta }}{\boldsymbol{t}}}$$
(3)

where, E is energy density, C represents specific capacitance, V is the cell voltage, P is power density, and Δt is the discharging time.

Figure 7d illustrates a Ragone plot (energy density vs power density) for the Fe3O4@CNFMn composite membrane supercapacitor and compares the currently fabricated supercapacitors with earlier reported works43,44,45,46,47,48,49,50,51,52. It could be noticed that the fabricated supercapacitors from Fe3O4@CNFMn demonstrated relatively high power density (~65 W/Kg) at high energy density (~13 Wh/kg) which is superior to many of the reported supercapacitors. This may be attributed to the stable electrolyte transport, superior surface area and relatively high number of micro- and mesopores, which provided a number of channels for easier ionic charge transport, thus improved the conductivity of the Fe3O4@CNFMn composite membrane and high power density.

## Discussion

Fe3O4@CNFMn composite nanofiber membrane has been successfully developed with bead on string structure by the combination of electrospinning and electrospray processes. The resultant membrane offered well scattered Fe3O4 particles within membrane matrix which turned the fragile nature of CNF into highly flexible structure, and MnO2 particles deposition throughout the membrane surface enabled fast and reversible faradic reaction and supplied shorter path for ionic diffusion. 3D micro-mesoporous structure of MnO2 provided larger contact area for electrode and electrolyte, supported higher electrolyte uptake, and also helped in scattering the external stress applied on the membrane to prevent membrane distortion. Quantitative performance analysis of synthesized nanofiber composite membrane showed that the membranes owed highly capacitive behavior which was apparent from rectangular shaped CV curves. Increase in current density decreased specific capacitance, whereas nonlinear behavior of charge discharge curves confirmed the contribution of pseudo capacitive mechanism. High conductivity and electrolyte uptake capacity owing to micro- and mesopores of the membrane leaded to an excellent specific capacitance of ~306 F/g. The high capacitance retention of ~85% even after 2000 cycles determined long and stable life of fabricated device. Flexibility of the synthesized membrane was further confirmed when fabricated supercapacitor was subjected to bending up to 180° and no significant change in specific capacitance was observed for up to 1000 bending cycles. Consequently, it was concluded that this technique offers a novel, valuable and scalable route for the designing flexible high-performance supercapacitors.

## Experimental Section

### Materials

Polyacrylonitrile (PAN, Mw = 90,000), polyvinyl alcohol (PVA, Mw = 86,000), Potassium hydroxide (KOH), Fe(acac)3, and KMnO4 were purchased from Aladdin Chemicals Co. Ltd., China. Dimethylformamide (DMF), and sulphuric acid (H2SO4), were provided by Shanghai Chemical Reagents Co., Ltd., China. All chemicals were used without further purification.

### Fabrication of the Fe3O4@CNF membrane

A homogenous solution comprising of Fe(acac)3 (1 wt.%) and PAN (8 wt.%) dissolved in DMF by magnetic stirring was electrospun following same conditions as reported in our earlier work18. Resultant precursor fibers were then vacuum dried and pre-oxidized at 280 °C for 2 h followed by carbonization at 800 °C for 2 h with a heating rate of 2 °C/min under nitrogen environment. Developed membrane showed ~50 µm thickness after treatment.

### Surface Modification and Electrospray of KMnO4

Prior to KMnO4 electrospray, Fe3O4@CNF composite membrane was rinsed with 0.1 M H2SO4 and washed with water then dried for 1 h at 50 °C. Typically, 5 wt.% dispersion of KMnO4 with PVA was electrosprayed directly onto subsequent surface modified Fe3O4@CNF using same electrospinning setup. A DC voltage of 15 kV was applied at a feeding rate of 1 mL/h and a distance of 20 cm between tip of the needles and Fe3O4@CNF (placed on stationary metallic cylinder) was maintained. Constant temperature and relative humidity were maintained at 23 ± 2 °C and 50 ± 5%, respectively. Later on, heat treatment was applied on resultant membrane to form MnO2 on the surface as shown in equation45.

$${\boldsymbol{Mn}}{{{\boldsymbol{O}}}_{{\bf{4}}}}^{-}+{\bf{4}}{{\boldsymbol{H}}}_{{\bf{2}}}{\boldsymbol{O}}\to {\boldsymbol{Mn}}{{{\boldsymbol{O}}}_{{\bf{2}}}}^{{\bf{2}}-}+{\bf{3}}{{\boldsymbol{O}}}_{{\bf{2}}}+{\bf{2}}{{\boldsymbol{H}}}_{{\bf{2}}}{\boldsymbol{O}}$$
(4)
$${\bf{4}}{\boldsymbol{Mn}}{{{\boldsymbol{O}}}_{{\bf{4}}}}^{-}+{\bf{3}}{\bf{C}}+{{\boldsymbol{H}}}_{{\bf{2}}}{\boldsymbol{O}}\to {\bf{4}}{\boldsymbol{Mn}}{{\boldsymbol{O}}}_{{\bf{2}}}+{\boldsymbol{C}}{{{\boldsymbol{O}}}_{{\bf{3}}}}^{{\bf{2}}-}+{\bf{2}}{\boldsymbol{HC}}{{{\boldsymbol{O}}}_{{\bf{3}}}}^{-}$$
(5)

### Structural Characterization

Field emission scanning electron microscope (FE-SEM, S-4800, Hitachi Ltd. Japan), high resolution transmission electron microscopy (HR-TEM, JEM-2100F, JEOL Ltd., Japan) and Raman spectroscopy system (invia-Reflex, Renishaw, Co., UK) were employed for morphological and structural analysis. X-ray diffraction (XRD) (D/Max-2550 PC Rigaku Co., Japan, Cu Kα, λ = 1.5406 Å) was involved for phase structural analysis and an automatic adsorption system (ASAP 2020, Micromeritics Co., USA) was used for examining Brunauer-Emmet-Teller (BET) surface area and porous structure by N2 adsorption-desorption isotherms. X-ray photoelectron spectroscopy (XPS) analysis was carried out on a PHI 5000 C ESCA system with Mg Kα source. Water contact angle (WCA) was tested by using a digital goniometer (Kino SL200B).

### Electrochemical Measurement

Electrochemical workstation (Chenhua CHI 660E, Shanghai) was employed for all electrochemical measurements. Two electrode setup was used to evaluate electrochemical performance of the all the samples. Prepared membranes were used directly employed as electrodes (each electrode weighed ~1 mg) to fabricate two electrode symmetric coin cells. Gel electrolyte (Na2SO4/PVA) was used as separator between two electrodes for all measurements. Moreover, in order to evaluate flexibility, each electrode was followed by gold sputtered polyethylene terephthalate (PET) on external side as protective layer. Cyclic voltammetry (CV) tests were carried out at various scan rates (i.e. 10–30 mV/s). Galvanostatic charge–discharge tests were analyzed using Chronopotentiometry (CP) at 1–8 A/g current density. Voltage drop of the synthesized device was calculated by charging at 0.5 V for 15 min followed by measuring the open circuit potential.

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

## References

1. 1.

Yan, J. et al. Recent advances in design and fabrication of electrochemical supercapacitors with high energy densities. Adv. Energy Mater. 4, 1300816 (2014).

2. 2.

Zhang, S. & Pan, N. Supercapacitors performance evaluation. Adv. Energy Mater. 5, 1401401 (2015).

3. 3.

Chen, X. et al. Smart, stretchable supercapacitors. Adv. Mater. 26, 4444–4449 (2014).

4. 4.

Gao, J. et al. Graphite-nanoplatelet-decorated polymer nanofiber with improved thermal, electrical, and mechanical properties. ACS Appl. Mater. Interfaces 5, 7758–7764 (2013).

5. 5.

Sun, H. et al. A novel slicing method for thin supercapacitors. Adv. Mater. 28, 6429–6435 (2016).

6. 6.

Brown, B. et al. Optimization of active manganese oxide electrodeposits using graphenated carbon nanotube electrodes for supercapacitors. Chem. Mater. 27, 2430–2438 (2015).

7. 7.

Ma, Y. et al. Graphene-based materials for lithium-ion hybrid supercapacitors. Adv. Mater. 27, 5296–5308 (2015).

8. 8.

Song, B. et al. Water-dispersible graphene/polyaniline composites for flexible micro-supercapacitors with high energy densities. Nano Energy 16, 470–478 (2015).

9. 9.

Zainab, G. et al. Free-standing, spider-web-like polyamide/carbon nanotube composite nanofibrous membrane impregnated with polyethyleneimine for CO2 capture. Compos. Commun. 6, 41–47 (2017).

10. 10.

Babar, A. A. et al. Tailoring differential moisture transfer performance of nonwoven/polyacrylonitrile-SiO2 nanofiber composite membranes. Adv. Mater. Interfaces 4, 1700062 (2017).

11. 11.

Iqbal, N. et al. Cobalt oxide nanoparticles embedded in flexible carbon nanofibers: attractive material for supercapacitor electrodes and CO2 adsorption. RSC Adv. 6, 52171–52179 (2016).

12. 12.

Kim, M., Lee, C. & Jang, J. Fabrication of highly flexible, scalable, and high-performance supercapacitors using polyaniline/reduced graphene oxide film with enhanced electrical conductivity and crystallinity. Adv. Funct. Mater. 24, 2489–2499 (2014).

13. 13.

Zhang, F. et al. Flexible films derived from electrospun carbon nanofibers incorporated with Co3O4 hollow nanoparticles as self-supported electrodes for electrochemical capacitors. Adv. Funct. Mater. 23, 3909–3915 (2013).

14. 14.

Iqbal, N. et al. Robust and flexible carbon nanofibers doped with amine functionalized carbon nanotubes for efficient CO2 capture. Adv. Sustainable Syst. 1, 1600028 (2017).

15. 15.

Ge, J. et al. Elastic and hierarchical porous carbon nanofibrous membranes incorporated with NiFe2O4 nanocrystals for highly efficient capacitive energy storage. Nanoscale 8, 2195–2204 (2016).

16. 16.

Lee, K. et al. Highly transparent and flexible supercapacitors using graphene-graphene quantum dots chelate. Nano Energy 26, 746–754 (2016).

17. 17.

Pan, Z. et al. Ultra-endurance flexible all-solid-state asymmetric supercapacitors based on three-dimensionally coated MnOx nanosheets on nanoporous current collectors. Nano Energy 26, 610–619 (2016).

18. 18.

Iqbal, N. et al. Highly flexible NiCo2O4/CNTs doped carbon nanofibers for CO2 adsorption and supercapacitor electrodes. J. Colloid Inter. Sci. 476, 87–93 (2016).

19. 19.

Iqbal, N. et al. In situ synthesis of carbon nanotube doped metal–organic frameworks for CO2 capture. RSC Adv. 6, 4382–4386 (2016).

20. 20.

Chen, T. & Dai, L. Flexible supercapacitors based on carbon nanomaterials. J. Mater. Chem. A 2, 10756–10775 (2014).

21. 21.

Liu, J. et al. Co3O4 Nanowire@MnO2 ultrathin nanosheet core/shell arrays: a new class of high-performance pseudocapacitive materials. Adv. Mater. 23, 2076–2081 (2011).

22. 22.

Xu, C. et al. Supercapacitive studies on amorphous MnO2 in mild solutions. J. Power Sources 184, 691–694 (2008).

23. 23.

Yang, Y. et al. Vertically aligned mixed V2O5-TiO2 nanotube arrays for supercapacitor applications. Chem. Commun. 47, 7746–7748 (2011).

24. 24.

Si, Y. et al. Fabrication of magnetic polybenzoxazine-based carbon nanofibers with Fe3O4 inclusions with a hierarchical porous structure for water treatment. Carbon 50, 5176–5185 (2012).

25. 25.

Guan, Q. et al. Needle-like Co3O4 anchored on the graphene with enhanced electrochemical performance for aqueous supercapacitors. ACS Appl. Mater. Interfaces 6, 7626–7632 (2014).

26. 26.

Mirvakili, M. et al. High-Performance Supercapacitors from Niobium Nanowire Yarns. ACS Appl. Mater. Interfaces 7, 13882–13888 (2015).

27. 27.

Yan, N. et al. Fe2O3 nanoparticles wrapped in multi-walled carbon nanotubes with enhanced lithium storage capability. Sci. Rep. 3, 3392 (2013).

28. 28.

Shi, C. et al. η-Fe2O3 spindles as anode materials for lithium-ion batteries with excellent stable and high electrochemical performance. Mater. Lett. 83, 35–38 (2012).

29. 29.

Goriparti, S. et al. Review on recent progress of nanostructured anode materials for Li-ion batteries. J. Power Sources 257, 421–443 (2014).

30. 30.

Kumar, P. et al. High capacity and low cost spinel Fe3O4 for the Na-ion battery negative electrode materials. Electrochim. Acta 146, 503–510 (2014).

31. 31.

Zhang, X. et al. Ultralong life lithium-ion battery anode with superior high-rate capability and excellent cyclic stability from mesoporous Fe2O3@TiO2 core–shell nanorods. J. Mater. Chem. A 2, 3912–3918 (2014).

32. 32.

Yang, Y. et al. Nanostructured amorphous MnO2 prepared by reaction of KMnO4 with triethanolamine. J. Alloys Compd. 505, 555–559 (2010).

33. 33.

Zhang, X. et al. Synthesis of a novel polyaniline-intercalated layered manganese oxide nanocomposite as electrode material for electrochemical capacitor. J. Power Sources 173, 1017–1023 (2007).

34. 34.

Tang, Y., Chen, T. & Yu, S. Morphology controlled synthesis of monodispersed manganese sulfide nanocrystals and their primary application in supercapacitors with high performances. Chemical Commun. 51, 9018–9021 (2015).

35. 35.

Wei, W. et al. Manganese oxide-based materials as electrochemical supercapacitor electrodes. Chem. Soc. Rev. 40, 1697–1721 (2011).

36. 36.

Gu, W. & Yushin, G. Review of nanostructured carbon materials for electrochemical capacitor applications: advantages and limitations of activated carbon, carbide-derived carbon, zeolite-templated carbon, carbon aerogels, carbon nanotubes, onion-like carbon, and graphene. Wiley Interdisciplinary Reviews: Energy Environ. 3, 424–473 (2014).

37. 37.

Li, J. et al. Doping effect of potassium permanganate on the performance of a copper/zinc oxide/alumina catalyst for methanol formation. Appl. Catal. A: General (1996) 139, 97–106 (1996).

38. 38.

Zhang, Y. et al. Influence of excess KMnO4 on the adsorption of powdered activated carbon. Chem. Eng. J. 226, 279–285 (2013).

39. 39.

Byles, B. W. et al. The role of electronic and ionic conductivities in the rate performance of tunnel structured manganese oxides in Li-ion batteries. APL Materials 4, 046108 (2016).

40. 40.

Luo, Q. et al. Effect of surface hydrophilicity on the supercapacitive performance of carbon paper. Ionics 23, 1915–1920 (2017).

41. 41.

Ren, T. et al. Polyacrylonitrile/polybenzoxazine-based Fe3O4@carbon nanofibers: hierarchical porous structure and magnetic adsorption property. J. Mater. Chem. 22, 15919–15927 (2012).

42. 42.

Fan, G. et al. Hierarchical porous carbon nanofibrous membranes with an enhanced shape memory property for effective adsorption of proteins. RSC Adv. 5, 64318–64325 (2015).

43. 43.

Ge, J. et al. Polybenzoxazine-based highly porous carbon nanofibrous membranes hybridized by tin oxide nanoclusters: durable mechanical elasticity and capacitive performance. J. Mater. Chem. A 4, 7795–7804 (2016).

44. 44.

Wang, G. et al. A review of electrode materials for electrochemical supercapacitors. Chem. Soc. Rev. 41, 797–828 (2012).

45. 45.

Le, T. et al. In-situ growth of MnO2 crystals under nanopore-constraint in carbon nanofibers and their electrochemical performance. Sci. Rep. 6, 37368 (2016).

46. 46.

Fan, R. et al. Facile synthesis and electrochemical properties ofα-MnO2as electrode material for supercapacitors. Asia-Pacific J. Chem. Eng. 8, 721–729 (2013).

47. 47.

Naoi, K. et al. Encapsulation of Nanodot Ruthenium Oxide into KB for Electrochemical Capacitors. J. Electrochem. Soc. 156, A52 (2007).

48. 48.

Reddy, M. et al. Multisegmented Au-MnO2/Carbon nanotube hybrid coaxial arrays for high-power supercapacitor applications. J. Phys. Chem. C 114, 658–663 (2010).

49. 49.

He, S. et al. Ultrathin MnO2 nanosheets supported on cellulose based carbon papers for high-power supercapacitors. J. Power Sources 246, 754–761 (2014).

50. 50.

Huang, Y. et al. High-performance flexible supercapacitors based on mesoporous carbon nanofibers/Co3O4/MnO2 hybrid electrodes. RSC Adv. 5, 18952–18959 (2015).

51. 51.

Chou, T. et al. Hierarchically porous carbon with manganese oxides as highly efficient electrode for asymmetric supercapacitors. ChemSusChem 7, 841–847 (2014).

52. 52.

Ko, W. et al. Porous honeycomb structures formed from interconnected MnO2 sheets on CNT-coated substrates for flexible all-solid-state supercapacitors. Sci. Rep. 6, 18887 (2016).

## Acknowledgements

This work is supported by the Program for Professor of Special Appointment (Eastern Scholar) at Shanghai Institutions of Higher Learning (No. TP2016019), the National Natural Science Foundation of China (Nos. 51503028 and 51673037), the Shanghai Rising-Star Program (No. 16QA1400200), the Shanghai Committee of Science and Technology (No. 15JC1400500), the National Key R&D Program of China (No. 2016YFB0303200), and the Fundamental Research Funds for the Central Universities.

## Author information

### Affiliations

1. #### State Key Laboratory for Modification of Chemical Fibers and Polymer Materials, College of Materials Science and Engineering, Donghua University, Shanghai, 201620, China

• Nousheen Iqbal
• , Xianfeng Wang
• , Aijaz Ahmed Babar
• , Ghazala Zainab
•  & Bin Ding
2. #### Key Laboratory of Textile Science & Technology, Ministry of Education, College of Textiles, Donghua University, Shanghai, 201620, China

• Nousheen Iqbal
• , Xianfeng Wang
•  & Bin Ding
3. #### Innovation Center for Textile Science and Technology, Donghua University, Shanghai, 200051, China

• Xianfeng Wang
• , Jianyong Yu
•  & Bin Ding

### Contributions

X.W., B.D., J.Y. and N.I. conceived and directed the experiments. N.I. and G.Z., A.A.B. performed the experiments and analyzed the data, N.I., X.W., A.A.B. and B.D. wrote the manuscript. All authors discussed the results and commented on the manuscript.

### Competing Interests

The authors declare that they have no competing interests.

### Corresponding authors

Correspondence to Xianfeng Wang or Bin Ding.

## Electronic supplementary material

### DOI

https://doi.org/10.1038/s41598-017-15535-x