Free standing hollow carbon nanofiber mats for supercapacitor electrodes

Abstract

Free standing hollow carbon nanofiber (CNF) mats with high graphitic content have been fabricated through co-axial electrospinning followed by high temperature pyrolysis. A blend of polyacrylonitrile (PAN) and poly(methyl methacrylate) (PMMA) in different weight ratios is used as the shell polymer whereas PMMA is used as the core polymer. The sacrificial PMMA template is removed during carbonization, creating a hollow core along with pores in the shell. In order to establish the best base case electrode, no further chemical or physical activation procedures or addition of metal oxide particles were employed. The structural and electrochemical properties of hollow CNFs as supercapacitor electrodes are systematically studied and compared with those of porous (PAN : PMMA weight ratios 2 : 1, 1 : 1, 1 : 2, 1 : 5) and solid CNFs. The nanofiber mats have been used directly as electrodes without binder and conductive additives owing to their good conductivity and mechanical stability. The hollow CNFs with the precursor polymer ratio of 1 : 5 (PAN : PMMA) in the shell, exhibited the highest specific surface area of 812.6 m² g⁻¹ with a large percentage of mesopores, delivering a capacitance of ∼185 F g⁻¹ at 5 mV s⁻¹, which was almost two orders of magnitude higher than the solid CNFs (1.2 F g⁻¹). In addition, the hollow CNFs exhibited an excellent charge/discharge capability delivering a capacitance of ∼105 F g⁻¹ at a current density of 2 A g⁻¹, with a capacitance retention of ∼80% after 3000 cycles. This study establishes the electrospun hollow CNFs as potential supercapacitor electrodes that can be easily modified further with the addition of functional materials.

---

1. Introduction

Electric double layer capacitors, owing to their high power densities, quick charge/discharge ability and longer cycle lives, are being considered a promising energy storage alternative to batteries and fuel cells.^1–5 The operation of an electric double layer capacitor (EDLC) is purely electrostatic in nature with energy being stored through charge separation in a nanometer thick double layer formed at the electrode–electrolyte interface.⁶ Since no transfer of electrons is involved across the interface, there are no charge transfer limitations like in batteries and hence a faster charge/discharge is possible.⁷ This feature makes EDLCs attractive over a wide range of applications including low power electronic devices to high power electric vehicles.

Carbon based materials are considered promising for energy storage and conversion devices as they are relatively low cost, easily processable, eco-friendly and can exist in a variety of forms with varied microtextures and dimensonalities.⁸ Various carbon materials like carbon aerogels,^9,10 activated carbons,^11–13 carbide derived carbon,^14–16 and carbon nanotubes^17–19 have been extensively examined as electrodes for double layer capacitors. Recently, focus has shifted to one dimensional carbon nanofibers (CNFs) which have emerged as promising electrode materials for EDLCs owing to their excellent physio-chemical properties like high specific surface area, very high aspect ratio and electrical conductivity offering efficient pathways for ion diffusion as well as chemical stability over a wide temperature and potential range in a variety of solutions.^20–25 Essentially, a high specific surface area with adequate pore size distribution and good electronic conductivity are the most important parameters which determine an EDLCs performance.²⁶ Usually, specific surface area of the carbons is increased by employing some physical or chemical activation method. The physical activation is mostly accomplished by treating the carbon material with steam or CO₂ whereas chemical activation is achieved through treatment with KOH, NaOH or ZnCl₂, thereby creating pores.^13,21,27 However, these activation techniques predominately create micropores (<2 nm) which do not add substantially to the electric double layer capacitance because of irreversible binding and the slow rate of electrolyte diffusion.¹⁹ Consequently, activated carbons show a huge capacitance drop at high current densities. For effective contribution to EDLC, minimum pore size should be >5 nm as validated through various studies.^28,29 Conventionally, mesopores are created in carbon materials using the template method. However, templating is complicated and time consuming process, as it requires a clean and complete removal of the template later in order to obtain the desired pore structure. In addition to the pore size effect, graphitic nature of the pore walls is also crucial in determining the capacitive behavior of the material.³⁰ It has been reported that the edge surfaces of the hexagon provide more effective double layer charging in comparison to the base plane surfaces.³¹ Hence a simple method for the fabrication of 1-D carbon materials having appropriate mesoporosity and graphitic structure is required for the material to be a promising supercapacitor electrode.

Electrospinning presents a simple route for fabricating fibers with diameters ranging from submicron to nanometer scales.^32–34 In electrospinning, a polymer melt is subjected to uniaxial elongation under the action of an electric field. When the applied electric field exceeds a threshold limit, electrostatic forces dominate over the surface tension force leading to distortion of the polymer drop into a Taylor cone and finally its ejection from the nozzle resulting into nanofibers.³⁵ Additional advantage with electrospinning is that nanofibers can be collected in the form of free-standing mats which post carbonization can be used directly as electrodes eliminating the need of binder and conductive additives.

In this study, hollow CNF mats have been fabricated by coaxial electrospinning using PAN and PMMA as precursors. PMMA is used in the core as sacrificial polymer which decomposes during carbonization. A polymeric emulsion of PAN and PMMA in different weight ratios is used as the shell fluid. The electrochemical performance of hollow, porous and solid CNFs is evaluated using cyclic voltammetry (CV), impedance spectroscopy and galvanostatic charge/discharge experiments. Specific capacitance of ∼105 F g⁻¹ is obtained at 2 A g⁻¹ suggesting hollow carbon nanofibers to be a simple base case electrode for electric double layer capacitors, which can be easily modified further with the addition of pseudocapacitive metal oxide nanoparticles. Although porous carbons synthesized by template method or by using polymer blends have been studied and evaluated in the past for their capacitive behavior.^36–39 However, to the best of our knowledge, this is the first study which is focused on the evaluation of PAN/PMMA derived hollow CNFs as freestanding supercapacitor electrodes.

2. Experimental

2.1. Fabrication

To fabricate hollow nanofibers, coaxial electrospinning was employed. A 24 wt% PMMA solution (molecular wt: 120 000, Sigma Aldrich) has been used as the sacrificial core polymer, whereas 12 wt% polymeric blend of PAN (molecular wt: 150 000, Sigma Aldrich) and PMMA in the ratio 2 : 1 and 1 : 5 has been used as shell fluid to fabricate the hollow fibers: H-21 and H-15 respectively. Both shell and core solutions are prepared using N,N-dimethylformamide (DMF) as the solvent. The solutions were ultrasonicated for 2 hours followed by stirring at 60 °C for 6 hours before electrospinning.

A 12 wt% polymeric blend of PAN : PMMA in the ratios 2 : 1, 1 : 1, 1 : 2 and 1 : 5 was electrospun to fabricate the porous CNFs: P-21, P-11, P-12 and P-15, respectively. The solid nanofibers were fabricated by electrospinning a 9 wt% PAN solution. For all the cases, a voltage of 15 kV was applied between the syringe needle and the rotating drum collector, whereas the distance between needle tip and the collector was maintained at 8 cm throughout the process. The nanofibers thus electrospun were collected in the form of mats and were oxidatively stabilized by heat treating them at 280 °C for 1 h in air. After stabilization, pyrolysis was carried out at 800 °C for 1 h in a controlled nitrogen atmosphere with heating rate maintained at 3 °C min⁻¹.

2.2. Characterization

The morphology and the internal microstructure of the CNFs were analyzed using a field emission scanning electron microscope, FESEM (Quanta 200, Zeiss, Germany) and a high-resolution transmission electron microscope, HRTEM (JEM-2100F JEOL, Japan), respectively. The BET surface area and the pore size of the fibers were evaluated from the N₂ adsorption–desorption isotherms recorded at 77 K using an Autosorb-1C machine (Quantachrome, USA). Raman spectroscopy (WiTec, Germany; k = 543 nm) was conducted to study the extent of graphitization in the hollow, porous and solid CNFs by calculating the ratio of graphitic to amorphous phases. Cyclic voltammetry (CV), galvanostatic charge/discharge and electrochemical impedance spectroscopy (EIS) were performed using a PGSTAT 302N electrochemical workstation (Metrohm Autolab, Netherlands). CVs were carried out at scan rates ranging from 5 mV s⁻¹ to 300 mV s⁻¹ within the potential range of −0.1 to 0 V in 6 M KOH solution using Pt wire as counter and Ag/AgCl as reference electrode. Galvanostatic charge/discharge measurements were done at three different current densities: 0.5 A g⁻¹, 1 A g⁻¹ and 2 A g⁻¹. Impedance measurements were done over the frequency range 1 mHz to 100 kHz with an a.c. signal of 10 mV amplitude. The electrical conductivity of the CNF mats was measured by four probe method using a Keithley 2602 digital electrometer.

3. Results and discussion

3.1. Physiochemical characterization

Fig. 1a–d show the FE-SEM micrographs of hollow CNFs: H-21 and H-15, respectively, having an average outer diameter of ∼300 nm and an approximate shell thickness of 30 nm. PMMA decomposes completely on thermal treatment (at 800 °C) creating a hollow core in the fiber along with pores in the shell. Fig. 1e–k show the surface morphology of the porous CNFs: P-11, P-12, P-15 and P-21 respectively. Fig. 1(l) shows the solid CNFs with a relatively smoother surface and an average diameter ∼300 nm. Pores are generated in the solid nanofibers by evaporation of the solvent. In the case of porous CNFs, the discontinuous PMMA phase in the polymeric blend burns out during carbonization creating channel/rod like pores. The internal microstructure of the hollow, porous and solid CNFs is clearly visible in the TEM images (Fig. 2).

Fig. 1 FE-SEM images of (a) hollow CNFs (H-21); (b) cross-section of a single hollow CNF (H-21); (c) hollow CNFs (H-15); (d) cross-sectional view of a single hollow CNF (H-15); (e and f) P-11 (PAN : PMMA = 1 : 1) porous CNFs; (g and h) P-12 (PAN : PMMA = 1 : 2) porous CNFs; (i and j) P-15 (PAN : PMMA = 1 : 5) porous CNFs; (k) P-21 (PAN : PMMA = 2 : 1) porous CNF; (l) solid CNFs.

Fig. 2 TEM images of (a) hollow CNF (PAN : PMMA ratio in shell = 2 : 1); (b) porous CNF (PAN : PMMA = 2 : 1); (c) porous CNF (PAN : PMMA = 1 : 5); and (d) solid CNFs.

Fig. 3 shows the HR-TEM micrographs of hollow CNFs (a) H-15, (b) H-21; porous CNFs (c) P-15, (d) P-12, (e) P-11, (f) P-21; and (g) solid CNFs. Hollow CNFs (H-15 and H-21) show some ordered graphitic domains near the shell walls of the nanofibers which agrees with the polymer chain orientation model.⁴⁰ It has been reported that as the fiber diameter is reduced (for e.g. thin shell region of hollow fibers), polymer chains align better inside the fiber due to the confinement effect, producing higher crystallinity.^41,42 The HR-TEM images of porous and solid CNFs mostly show a turbostratic carbon structure, with large stacking disorder.

Fig. 3 HRTEM lattice fringe images of the CNFs depicting their graphitic structure: (a) hollow (H-15) CNF showing higher graphitization near the shell walls; (b) hollow (H-21) CNF showing higher graphitization near the shell walls; (c) porous CNF (P-15); (d) porous CNF (P-12); (e) porous CNF (P-11); (f) porous CNF (P-21); (g) solid CNF, showing a turbostratic carbon structure.

The evolution of porosity and its effect on the adsorption behavior of the CNFs was examined by N₂ adsorption at 77 K. The BET specific surface area and pore-size distribution of hollow (H-15 and H-21), porous (P-15, P-12 and P-21) and solid CNFs are summarized in Table 1. Fig. 4 shows the adsorption/desorption isotherms of the CNFs. The isotherms of H-15, H-21, P-15, P-12 and P-21 CNFs can be approximated as type IV having slight hysteresis loops, indicative of capillary condensation in macro and mesopores.⁴³ Nitrogen adsorption in the low-pressure region is associated with the filling of micropores whereas adsorption in the high pressure region is due to the mesopores which are created mainly due to the burning out of PMMA during thermal treatment. Reduced density of micropores in the hollow and porous fibers is probably due to the coalescence of nearby micropores into mesopores and macropores.

Table 1 Specific surface area and pore size distribution for solid, porous and hollow CNFs

CNF	BET S.S.A (m^2 g^−1)	Avg. pore dia (nm)	VT (cm^3 g^−1)	Vmeso (cm^3 g^−1)	Vmicro (cm^3 g^−1)
Solid	41.7	2.01	0.021	0.0121	0.0089
Porous (2 : 1), P-21	416.8	4.68	0.4877	0.3301	0.1576
Porous (1 : 1), P-11	460.2	4.72	0.4956	0.3374	0.1582
Porous (1 : 2), P-12	568.5	5.23	0.8926	0.6120	0.2806
Porous (1 : 5), P-15	695.2	5.76	0.9127	0.6331	0.2796
Hollow (2 : 1), H-21	566.5	6.22	0.8814	0.6047	0.2767
Hollow (1 : 5), H-15	812.6	6.31	0.9464	0.6642	0.2822

---

Fig. 4 Nitrogen adsorption–desorption isotherm of H-15, H-21, P-15, P-12, P-21, and solid CNFs.

Fig. 5 shows the Raman spectra of solid, porous (P-15, P-12, P-11 and P-21) and hollow (H-15 and H-21) CNFs. The peaks at 1352 cm⁻¹ and 1590 cm⁻¹ represent the D band and G bands, respectively. The ratio of the intensities of D and G bands, R = (I_D/I_G) which indicates the degree of graphitization in a carbon material were calculated to be 1.44, 1.34, 1.26, 1.22, 1.15, 1.21 and 1.12 for the solid, porous (P21, P11, P12, P15) and hollow (H21, H15) CNFs, respectively. The R values were obtained by fitting the Gaussian–Lorentzian function to the Raman spectra after deconvolution. The hollow CNF, H-15 shows the highest graphitization as is indicated by its R value, which is lowest. High percentage of graphitic domains in H-15 can be attributed to its high surface area exposure during pyrolysis, as graphitization occurs more on the surface compared to the bulk material.⁴⁴ Further, the better graphitic structure observed in hollow CNFs is can be the result of an improved macromolecular alignment in the thinner shell region (∼30 nm) of the precursor fibers that translates to an improved graphitic structure and crystal orientation during pyrolysis. The thinner fiber diameter (shell region of the hollow fiber) results in an increased polymer chains alignment due to the confinement effect, producing higher crystallinity.⁴¹

Fig. 5 Raman spectra of solid, porous (P-15, P-12, P-11, P-21) and hollow (H-15, H-21) CNFs.

3.2. Electrochemical characterization

3.2.1. Cyclic voltammetry. The electrochemical behavior of the CNF mats was investigated by cyclic voltammetry in a three electrode cell. Fig. 6a–g show the cyclic voltammograms of H-15, P-15, P-12, P-11, H-21, and P-21 and solid nanofiber mat electrodes at scan rates ranging from 5 mV s⁻¹ to 300 mV s⁻¹. For all the CNF electrodes (solid, porous and hollow), shape of the CV curves obtained is almost rectangular with no faradaic peaks observed, indicating the capacitance obtained is purely due to the double layer charging and no charge transfer has taken place.

Fig. 6 Cyclic voltammogram of (a) hollow CNF (H-15), (b) porous CNF (P-15), (c) porous CNF (P-12), (d) porous CNF (P-11), (e) hollow CNF (H-21), (f) porous CNF (P-21) and (g) solid CNF electrodes at different scan rates, (h) specific capacitance vs. scan rate for the solid, porous (P-15, P-12, P-11 and P-21) and hollow (H-21 and H-15) CNF electrodes.

Specific capacitances are calculated from the CV curves by using the equation,

C (F g⁻¹) = I/m × (dV/dt),

where I (Amp) is the average current, dV/dt (V s⁻¹) is the applied scan rate and m (g) is the weight of the nanofiber mat. The hollow CNF, H-15 produced the highest specific capacitance, 185 F g⁻¹ at 5 mV s⁻¹ (Fig. 6a). It can be observed in Fig. 6a that a nearly rectangular CV curve is maintained even at the very high scan rate of 300 mV s⁻¹ with a specific capacitance of 105 F g⁻¹, implying that the electrode H-15 has an excellent rate capability with ability to undergo charge/discharge very quickly. The porous CNFs: P-15, P-12, P-11 and P-21 exhibit comparatively lower specific capacitances of 145.5, 132.6, 61.3 and 38.5 F g⁻¹ respectively. Solid CNFs on the other hand, exhibit very low specific capacitance of 1.2 F g⁻¹ at 5 mV s⁻¹. Fig. 6h shows the specific capacitances calculated at different scan rates for the hollow (H-15 and H-21), porous (P-15, P-12, P-11 and P-21) and solid CNFs.

The capacitive behavior of the CNFs can be directly correlated with their specific surface area and pore diameters. The better electrochemical performance of H-15 CNF compared to the solid and porous CNFs can be attributed to its high BET surface area (∼812 m² g⁻¹) having a large mesopore fraction (∼70.18%), and an appropriate pore diameter (∼6.31 nm). Among the porous electrodes, P-15 shows the highest specific capacitance of 145 F g⁻¹, mainly due to its higher specific surface area compared to the rest of porous CNFs. As the average pore size is comparable with the size of the electrolyte ion, ions can enter and leave the hollow fibers quickly facilitating quick charge/discharge. Higher percentage of mesopores in the hollow fibers help increase the ion accessible surface, facilitating fast motion of electrolyte ions in the pores even at high scan rates. The core of the hollow nanofiber can provide more space to accumulate electrolyte ions, thus enhancing the capacitance.

3.2.2. Galvanostatic charge/discharge. Fig. 7a and c show the charge/discharge voltage profiles of the hollow (H-15) and porous (P-15) CNF electrodes at three different current densities 0.5, 1 and 2 A g⁻¹. The specific capacitance is calculated from the charge–discharges curves using the following equation,

C = It/mΔV,

where, C is the gravimetric capacitance of the electrode (F g⁻¹), I is the discharge current (A), t is discharge time (s), ΔV is the voltage drop during the time of discharge, and m (g) is weight of the active material. In this study, the CNF mat has been used directly as a free-standing electrode without any binder, hence weight of the active material is same as the weight of the mat. The hollow H-15 electrode exhibits specific capacitance of 192, 126 and 104 F g⁻¹ at current densities 0.5, 1 and 2 A g⁻¹ which is comparable with the results of the CV measurements. The porous P-15 electrode shows specific capacitance of 141, 111 and 72 F g⁻¹ at current densities 0.5, 1 and 2 A g⁻¹, which also conforms with their CV results. The charge–discharge profiles are almost linear over the entire voltage range confirming that the capacity observed is due to double layer build-up only and not due to side reactions. The electrochemical stability of H-15 and P-15 electrodes was evaluated by running the galvanostatic charge–discharge test for 3000 cycles. Fig. 7c and d show the specific capacitance as a function of the cycle number for H-15 and P-15 electrodes. For the H-15 electrode, it is observed that the specific capacitance has decreased only by 5%, 11% and 19% at current densities 0.5, 1 and 2 A g⁻¹, respectively, after 3000 cycles. For the P-15 electrode, the decrease in specific capacitance after 3000 cycles is calculated to be ∼5%, 12% and 15% at current densities 0.5, 1 and 2 A g⁻¹ respectively. The results reveal that both H-15 and P-15 electrodes exhibit long term electrochemical stability, and can be established as the best base case supercapacitor electrodes whose performance can be further enhanced by addition of higher capacitance metal oxide nanoparticles.

Fig. 7 Galvanostatic charge–discharge curve at current densities 0.5, 1 and 2 A g⁻¹ of (a) H-15 CNF electrode, (c) P-15 CNF electrode. Variation in specific capacitance as a function of cycle number of (b) H-15 CNF electrode, (d) P-15 CNF electrode.

Fig. 8a and b show the FE-SEM images of the nanofiber mat electrodes, H-15 and P-15 at the end of 3000 cycles. It can observed in the images that the fiber structure and morphology is intact even after continued cycling, owing to the intertwined web structure of nanofibers that imparts good mechanical stability to the mat. Fig. 8c and d show the FE-SEM image of the H-15 free standing mat electrode before electrochemical cycling and after 3000 charge–discharge cycle. In order to further confirm the structural stability of the nanofibers, TEM images were taken after the completion of cycling test. For this, a small portion of the nanofibers mat was crushed and dispersed in ethanol, followed by drop casting on a 300 mesh carbon coated copper grid for visualization. Fig. 9a–d show the TEM images H-15 and P-15 CNFs, respectively, with their structures still maintained after 3000 cycles.

Fig. 8 Morphology of the CNF mats after 3000 cycles (a) hollow (H-15), (b) porous (P-15) CNF mat electrode. Lower magnification image of the free standing CNF mat (c) before electrochemical cycling and (d) after 3000 charge–discharge cycles.

Fig. 9 TEM images of the CNFs at the end of 3000 cycles demonstrating their structural stability (a), (b) and (c) H-15; (d) P-15.

Further, energy and power densities were calculated for the hollow CNF by using the following equations:

Energy density, E (W h kg⁻¹) = (½) × C × V² × 1000 (g kg⁻¹) × 1/3600 (W h J⁻¹)

Power density, P (W kg⁻¹) = E/t

The specific capacitances, energy and power densities of H-15 and P-15 CNF electrodes at current densities 0.5, 1 and 2 A g⁻¹ are summarized in Table 2.

Table 2 Specific capacitance, energy and power densities of H-15 and P-15 CNF electrodes at different current densities

CNF electrode	Current density (A g^−1)	Sp. capacitance (F g^−1)	Energy density (W h kg^−1)	Power density (W kg^−1)
H-15	0.5	192	26.7	249.9
	1	126	17.5	500
	2	104	14.4	996.9
P-15	0.5	141	19.58	250
	1	111	15.4	499.4
	2	72	10	1000

---

3.2.3. Impedance spectroscopy. Further, impedance spectroscopy was performed to study the electrochemical behavior of the electrodes. Fig. 10a shows the Nyquist plot of the hollow (H-15 and H-21), porous (P-15, P-12, P-11 and P-21) and solid CNFs. The impedance behavior of electrodes can be represented by a network of circuit elements consisting of equivalent series resistance (ESR), double layer capacitance, faradaic charge transfer resistance and the Warburg impedance element.¹ The high frequency region at the beginning of the graph represents ESR which comprises of the solution resistance, resistance at the electrolyte/electrode interface and the contact resistance between the collector and the electrode. The values of ESR obtained from the intercept on the real axis are 0.05, 0.3, 0.57, 0.65, 0.6, 0.85 and 2.8 Ohms for H-15, P-15, P-12, P-11, H-21, P-21 and solid CNF electrodes, respectively. The straight line in the low frequency region represents the electrolyte diffusion into the bulk of the electrode, also known as the Warburg impedance. In the case of the hollow fibers, the points in the Warburg zone are almost dispersed vertically up representing good capacitive behavior. In the case of solid and porous fibers, however the Warburg region is not a vertical line, which is also reflected in the CV curves in terms of their low capacitive values. The superior performance of the hollow fiber (H-15) can be ascribed to its higher electrical conductivity and enlarged pore size.

Fig. 10 (a) Nyquist plot of hollow (H-15, H-21), porous (P-15, P-12, P-11, P-21) and solid CNF electrodes ac signal amplitude is 10 mV with frequency range from 10 mHz to 100 kHz. (b) Electrical conductivities of H-15, P-15 and solid CNF measured by four-probe method.

The electrical conductivities of the CNF mat electrodes were measured by four point probe. The electrical conductivity values of H-15, H-21, P-15, P-12, P-11, P-21 and solid CNF electrodes were calculated to be 1.25 × 10³, 1.03 × 10³, 4.5 × 10², 4.1 × 10², 1.38 × 10², 1.31 × 10² and 3 × 10 S m⁻¹, respectively. Fig. 10b shows the current verses potential plot of H-15, P-15 and solid nanofiber mats. The electrical resistivity values were calculated by using the equation ρ = 4.52 × (V/I) × t, where V, I, t and ρ denote the measured voltage (V), current (A), thickness (m) of the mat and resistivity (Ω m), respectively.⁴⁵ The reciprocal of electrical resistivity gives the electrical conductivity by using the relation σ = 1/ρ, where σ denotes the electrical conductivity (S m⁻¹). The higher electrical conductivity of H-15 CNF which resulted in faster charge transfer at the electrode/electrolyte interface is also validated by the EIS spectra. The higher electrical conductivity observed in H-15 electrode can be attributed to: (a) its higher surface area which led to a higher graphitization at the surface during pyrolysis, (b) improved macromolecular alignment in the thin shell region of precursor fibers that translated into a better graphitic structure on pyrolysis. A high electrical conductivity is desired in supercapacitors for minimizing the internal resistance in the cell, and to facilitate a faster transfer of the applied electrical potential to the electrode/electrolyte interface.

4. Conclusions

Standalone binder-free carbon electrodes were obtained by carbonization of electrospun polymeric nanofibers. The CNF mats could be used directly as supercapacitor electrodes without adding binder and conducting additives owing to their electrical conductivity and mechanical robustness. Compared to the porous and solid CNFs, hollow CNF exhibited a high specific surface area with a large volume fraction of mesopores. The hollow core provided a large accessible contact area facilitating quick adsorption and desorption of electrolyte ions resulting in almost ideal capacitive behavior even at fast scan rates. Specific capacitance of 185 F g⁻¹ is obtained for the hollow (H-15) CNFs at 5 mV s⁻¹ which is significantly higher than the porous CNFs and almost two orders of magnitude higher than that of solid CNFs. Further, the H-15 electrode was able to deliver a high power density while maintaining a sufficiently high energy density. The Nyquist plot showed that the capacitor cell with hollow CNF as electrode has a very low ESR and an almost vertical line for Warburg impedance proving hollow electrospun CNFs to be a very promising base material for EDLC applications.

Acknowledgements

This work was supported by the Department of Science and Technology, New Delhi through its Nanoscience Center at IITK.

References

1. B. E. Conway, Electrochemical Supercapacitors: Scientific Fundamentals and Technological Applications, Kluwer Academic Publisher/Prenum Press, New York, 1999.
2. R. Kotz and M. Carlen, Electrochim. Acta, 2000, 45, 2483–2498.
3. M. Endo, T. Maeda, T. Takeda, J. Kim, K. Koshiba, H. Hara and M. S. Dresselhaus, J. Electrochem. Soc., 2001, 148, A910–A914.
4. K. Kinoshita, Carbon: Electrochemical and Physicochemical Properties, John Wiley & Sons, Inc., New York, 1996.
5. J. R. Miller and P. Simon, Science, 2008, 321, 651–652.
6. P. Simon and Y. Gogotsi, Nat. Mater., 2008, 7, 845–854.
7. M. Winter and R. J. Brodd, Chem. Rev., 2004, 104, 4245–4269.
8. E. Frackowiak and F. Béguin, Carbon, 2001, 39, 937–950.
9. D. Carriazo, F. Picó, M. C. Gutiérrez, F. Rubio, J. M. Rojo and F. del Monte, J. Mater. Chem., 2010, 20, 773–780.
10. J. Li, X. Wang, Q. Huang, S. Gamboa and P. J. Sebastian, J. Power Sources, 2006, 158, 784–788.
11. O. Barbieri, M. Hahn, A. Herzog and R. Kötz, Carbon, 2005, 43, 1303–1310.
12. E. Taer, M. Deraman, I. A. Talib, A. Awitdrus, S. A. Hashmi and A. A. Umar, Int. J. Electrochem. Sci., 2011, 6, 3301–3315.
13. C. Kim, J. Power Sources, 2005, 142, 382–388.
14. Y. Gogotsi, A. Nikitin, H. Ye, W. Zhou, J. E. Fischer, B. Yi, H. C. Foley and M. W. Barsoum, Nat. Mater., 2003, 2, 591–594.
15. J. Chmiola, C. Largeot, P. L. Taberna, P. Simon and Y. Gogotsi, Science, 2010, 328, 480–483.
16. Y. Korenblit, M. Rose, E. Kockrick, L. Borchardt, A. Kvit, S. Kaskel and G. Yushin, ACS Nano, 2010, 4, 1337–1344.
17. B. Kim, H. Chung and W. Kim, Nanotechnology, 2012, 23, 155401.
18. E. Frackowiak, K. Metenier, V. Bertagna and F. Beguin, Appl. Phys. Lett., 2000, 77, 2421–2423.
19. K. H. An, W. S. Kim, Y. S. Park, J. M. Moon, D. J. Bae, S. C. Lim, Y. S. Lee and Y. H. Lee, Adv. Funct. Mater., 2001, 11, 387–392.
20. C. Kim and K. S. Yang, Appl. Phys. Lett., 2003, 83, 1216–1218.
21. E. J. Ra, E. Raymundo-Piñero, Y. H. Lee and F. Béguin, Carbon, 2009, 47, 2984–2992.
22. Y. Yang, A. Centrone, L. Chen, F. Simeon, T. A. Hatton and G. C. Rutledge, Carbon, 2011, 49, 3395–3403.
23. C.-W. Huang, Y.-T. Wu, C.-C. Hu and Y.-Y. Li, J. Power Sources, 2007, 172, 460–467.
24. C. Kim, B. T. N. Ngoc, K. S. Yang, M. Kojima, Y. A. Kim, Y. J. Kim, M. Endo and S. C. Yang, Adv. Mater., 2007, 19, 2341–2346.
25. Y. E. Miao, W. Fan, D. Chen and T. Liu, ACS Appl. Mater. Interfaces, 2013, 5, 4423–4428.
26. L. L. Zhang and X. S. Zhao, Chem. Soc. Rev., 2009, 38, 2520–2531.
27. P. Li, T.-J. Zhao, J.-H. Zhou, Z.-J. Sui, Y.-C. Dai and W.-K. Yuan, Carbon, 2005, 43, 2701–2710.
28. S. T. Mayer, R. W. Pekala and J. L. Kaschmitter, J. Electrochem. Soc., 1993, 140, 446–451.
29. I. Tanahashi, A. Yoshida and A. Nishino, J. Electrochem. Soc., 1990, 137, 3052–3057.
30. S. Mitani, S.-I. Lee, K. Saito, Y. Korai and I. Mochida, Electrochim. Acta, 2006, 51, 5487–5493.
31. J.-P. Randin and E. Yeager, J. Electroanal. Chem. Interfacial Electrochem., 1972, 36, 257–276.
32. M. Inagaki, Y. Yang and F. Kang, Adv. Mater., 2012, 24, 2547–2566.
33. D. Li and Y. Xia, Adv. Mater., 2004, 16, 1151–1170.
34. I. G. Loscertales, A. Barrero, M. Márquez, R. Spretz, R. Velarde-Ortiz and G. Larsen, J. Am. Chem. Soc., 2004, 126, 5376–5377.
35. D. H. Reneker, A. L. Yarin, H. Fong and S. Koombhongse, J. Appl. Phys., 2000, 87, 4531–4547.
36. Z. Ma, T. Kyotani and A. Tomita, Carbon, 2002, 40, 2367–2374.
37. C. Vix-Guterl, S. Saadallah, K. Jurewicz, E. Frackowiak, M. Reda, J. Parmentier, J. Patarina and F. Beguin, Mater. Sci. Eng., 2004, 108, 148–155.
38. X. Zhang, X. Wang, Y. Bai, X. Wang and J. Su, J. Electrochem. Soc., 2012, 159, A431–A437.
39. L. Ji, Z. Lin, A. J. Medford and X. Zhang, Carbon, 2009, 47, 3346–3354.
40. Y. Ji, C. Li, G. Wang, J. Koo, S. Ge, B. Li, J. Jiang, B. Herzberg, T. Klein, S. Chen, J. C. Sokolov and M. Rafailovich, EPL, 2008, 84, 1–6.
41. D. Papkov, A. M. Beese, A. Goponenko, Y. Zou, M. Naraghi, H. D. Espinosa, B. Saha, G. C. Schatz, A. Moravsky, R. Loutfy, S. T. Nguyen and Y. Dzenis, ACS Nano, 2013, 7, 126–142.
42. J. C. Jenkins and G. M. Jenkins, Fuel, 1987, 66, 1519–1521.
43. C. Kim, Y. Jeong II, B. T. N. Ngoc, K. S. Yang, M. Kojima, Y. A. Kim, M. Endo and J.-W. Lee, Small, 2007, 3, 91–95.
44. Y. Tanabe, J. Yamanaka, K. Hoshi, H. Migita and E. Yasuda, Carbon, 2001, 39, 2347–2353.
45. R. Kumar, K. Jahan, R. K. Nagarale and A. Sharma, ACS Appl. Mater. Interfaces, 2015, 7, 593–601.

---