- Email: [email protected]
Optimization of the pore structure of PAN-based carbon fibers for enhanced supercapacitor performances via electrospinning
Accepted Manuscript Optimization of the pore structure of PAN-based carbon fibers for enhanced supercapacitor performances via electrospinning Young-Jung Heo, Hyo In Lee, Ji Won Lee, Mira Park, Kyong Yop Rhee, Soo-Jin Park PII:
S1359-8368(18)32525-3
DOI:
10.1016/j.compositesb.2018.10.026
Reference:
JCOMB 6104
To appear in:
Composites Part B
Received Date: 9 August 2018 Revised Date:
7 October 2018
Accepted Date: 9 October 2018
Please cite this article as: Heo Y-J, Lee HI, Lee JW, Park M, Rhee KY, Park S-J, Optimization of the pore structure of PAN-based carbon fibers for enhanced supercapacitor performances via electrospinning, Composites Part B (2018), doi: https://doi.org/10.1016/j.compositesb.2018.10.026. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
1
Manuscript submitted to “Composites Part B: Engineering” as an original paper
2
RI PT
3
4
Optimization of the pore structure of PAN-based carbon fibers for
6
enhanced supercapacitor performances via electrospinning
M AN U
SC
5
7
Young-Jung Heoa, Hyo In Leea, Ji Won Lee, Mira Parkb, Kyong Yop Rheec,** and Soo-
9
Jin Parka,*
a
10
b
12
17
EP
c
Department of Mechanical Engineering, College of Engineering, Kyung Hee University,
AC C
16
Department of Bioenvironmental Chemistry, College of Agriculture & Life Science, Chonbuk National University, Jeonju 54896, South Korea
13
15
Department of Chemistry, Inha University, 100 Inharo, Incheon 22212, Republic of Korea
11
14
TE D
8
Yongin, 17104, Republic of Korea
18
*,** Corresponding authors.
19
Tel.: +82-32-876-7234; Fax: +82-32-867-5604.
20
E-mail addresses : [email protected] (K.Y. Rhee), [email protected] (S.-J. Park).
21 1
ACCEPTED MANUSCRIPT
22
Abstract
Activated microporous polyacrylonitrile-based carbon nanofibers (APCFs)
24
were synthesized by a sequential process of electrospinning, carbonization, and KOH
25
activation. The porosity and surface chemistry of the APCFs strongly depended on the
26
activation temperature. The specific surface area and pore volume varied from 15 to
27
1886 m2 g-1 and 0.021 to 1.196 cm3 g-1, respectively, as the activation temperature
28
increased; this was accompanied by morphology changes at high temperature. The
29
dominant microstructure and minor mesostructure improved the capacitance of carbon.
30
Compared to the other samples, APCFs activated at an optimum temperature of 1000 °C
31
showed the highest specific capacitance of 103.01 F g-1 at 1 A g-1 in 1 mol L-1 Na2SO4
32
aqueous electrolyte, and an excellent cycling durability up to 3000 cycles. The
33
improved electrochemical efficiency could be explained by the high specific surface
34
area, suitable pore size, and influence of heteroatoms relative to the increased electrical
35
double-layers. The change in the pore size distribution with activation temperature is
36
also discussed in detail.
37
AC C
EP
TE D
M AN U
SC
RI PT
23
Keywords: Supercapacitor, Polyacrylonitrile, Porous carbon fibers, KOH activation
38 2
ACCEPTED MANUSCRIPT
39
1. Introduction
Fossil fuels are fast depleting owing to the rapidly growing energy demands,
41
and much study has been dedicated to developing high-efficiency, alternative, and
42
environment friendly energy storage devices such as supercapacitors, fuel cells, and
43
lithium-ion batteries [1-3].
SC
RI PT
40
Supercapacitors, also known as ultracapacitors, have emerged as the most
45
promising energy storage devices for portable electronics and hybrid electric vehicles
46
because of their high power density, long cycle life, and fast charge-discharge rates [4,
47
5]. They have additional advantages such as low maintenance costs and safe operation
48
[6]. Based on the charge-discharge mechanisms, supercapacitors can generally be
49
divided
50
pseudocapacitors. Charge-discharge of EDLCs occurs at the electrode/electrolyte
51
interface by reversible adsorption of electrolyte ions (non-faradaic reactions and
52
physical adsorption). Pseudocapacitors store energy by reversible faradaic redox
53
reactions (chemical) between the active materials and the electrolyte [7, 8].
54
Pseudocapacitors exhibit high capacitance (nearly a dozen times higher than that of
55
EDLCs); however, they are not suitable for industrial applications due to their poor
two
types:
electrical
double-layer
capacitors
(EDLCs)
and
AC C
EP
into
TE D
M AN U
44
3
ACCEPTED MANUSCRIPT
cycle life and expensive active materials used for the production (usually metal oxides
57
such as RuO2, IrO2, and MnO2) [9]. Thus, research has focused on developing
58
commercial supercapacitors based on carbon electrode materials because of their low
59
cost and high stability, despite their low energy density [10]. Various kinds of carbon
60
materials, such as carbon nanofibers, graphene oxide, carbon nanotubes, and activated
61
carbon, have been used as electrode materials [11]. Carbon nanofibers are employed as
62
supercapacitors as they possess excellent chemical stability, good cyclability, and high
63
conductivity.
64
Currently, many carbon materials are produced from polymer precursors, e.g., cellulose-
65
based fibers, polyacrylonitrile (PAN)-based fibers, and pitch-based fibers [12]. PAN, as
66
a precursor of carbon fibers, exhibits a high carbon yield after carbonization and
67
possesses superior mechanical properties [13]. There are several PAN-based fiber
68
preparation techniques, of which the electrospinning technique, which uses electrostatic
69
forces to prepare the fibers, is the most widely used [11, 13]. Electrospinning is a simple
70
and useful method when there are multiple components in the precursor solution, and
71
can produce uniform size-ordered fibers in the nano- to micrometer diameter range.
72
PAN is characterized by good spinnability, which makes it generally suitable as an
AC C
EP
TE D
M AN U
SC
RI PT
56
4
ACCEPTED MANUSCRIPT
electrospinning material. However, pristine PAN-based fibers present poor capacitance
74
owing to their low specific surface area. The specific surface area and porosity are the
75
most important factors for enhanced capacitance, because energy storage in a carbon
76
electrode is grounded by the electrostatic adsorption of electrolyte ions on the large
77
interface between the electrode and the electrolyte [14]. Chemical activation is usually
78
used to increase the specific surface area, for which KOH is the most powerful and
79
accessible activation agent [15]. In pore size analysis of carbon materials, Horvath-
80
Kawazoe (HK) and Barret-Joyner-Hallenda (BJH) methods have been widely used for
81
micropore and mesopore analysis on porous carbons, respectively [16,17]. However,
82
their methods have limitations due to their limited pore range and poor accuracy. In the
83
case of the Density Functional Theory (DFT) method, it has the advantage of being able
84
to analyze from micro to meso size simultaneously with more accurate analysis result
85
than previous methods [18,19].
SC
M AN U
TE D
EP
AC C
86
RI PT
73
In this study, PAN-based carbon fibers (PCFs) were synthesized from PAN
87
powder via electrospinning, and activated at various temperatures with KOH, to prepare
88
activated PAN-based carbon fibers (APCFs). The obtained highly microporous APCFs
89
were used as electrodes in EDLCs, and their electrochemical performance was 5
ACCEPTED MANUSCRIPT
evaluated to optimize the porous structure and thus obtain good electrochemical
91
efficiency.
92
2. Experimental
93
2.1. Materials
SC
RI PT
90
For the electrospinning solution, N,N-dimethylformamide (DMF, ≥99.5%,
95
Daejung Co.) and polyacrylonitrile (PAN, Mw = 150,000, Sigma-Aldrich Co.) were
96
used. KOH (Duksan Pure Chemicals Co.) and HCl (35%, Samchun Pure Chemical Co.)
97
were used as the chemical activation agents. The reagents used to prepare the slurry for
98
the electrochemical measurements were polyvinylidene fluoride (PVDF, Mw = 534,000,
99
Sigma-Aldrich Co.) and 1-methyl-2-pyrrolidone (NMP, ≥99.0%, TCI Co.). All
TE D
M AN U
94
chemicals were used without further purification.
101
2.2. Preparation of PAN-based carbon fibers
AC C
102
EP
100
The PAN solution for electrospinning was prepared by dissolving 10 wt% PAN
103
in DMF by stirring for 12 h at room temperature. The prepared polymer solution was
104
loaded into a 12 mL syringe and ejected from a stainless-steel needle onto an aluminum
105
foil-covered cylindrical collector using an electrospinning apparatus (Nano NC). The 6
ACCEPTED MANUSCRIPT
electrospinning equipment utilized a supplied voltage of 15 kV, feeding rate of 0.8 mL
107
h-1, tip-to-collector distance of 15 cm, and collector speed of 280 rpm, with a syringe
108
pump (KDS-200, KD Scientific). The temperature and relative humidity during
109
spinning process were maintained at 25±2 °C and 50±5%, respectively. The PAN fibers
110
were stabilized at 280 °C for 1 h under air and subsequently carbonized at 1000 °C for 3
111
h under N2.
112
2.3. Chemical activation using potassium hydroxide
M AN U
SC
RI PT
106
The pristine PAN-based carbon fibers (PPCFs) were impregnated with a
114
mixture of distilled water (20 mL), ethanol (5 mL), and KOH (weight ratio of
115
KOH:PPCFs = 4:1), and subsequently dried to a constant weight at 80 °C. The soaked
116
PPCFs were activated at various temperatures (600–1000 °C) for 1 h under N2
117
atmosphere, and labelled as APCF-600, APCF-800, APCF-900, and APCF-1000.
118
Activation temperatures above 1000 °C led to poor yields (≤5%) of APCF. After the
119
activation, the obtained APCFs were first washed with HCl (0.5 M) for neutralization
120
and removal of residual potassium atoms or ions, following which they were filtered
121
with distilled water until the pH of the filtrate reached approximately 6–7. The washed
122
APCFs were dried overnight at 60 °C. A schematic of the PPCF and APCF preparation
123
processes is presented in Fig. 1.
AC C
EP
TE D
113
7
ACCEPTED MANUSCRIPT
124
2.4. Characterization The morphologies of the samples were evaluated using scanning electron
126
microscopy (SEM, Hitachi, SU8010) and transmission electron microscopy (TEM,
127
Philips, CM200). The chemical bond energy on the surfaces and elemental composition
128
were investigated by X-ray photoelectron spectroscopy (XPS, Thermo Scientific, K-
129
Alpha). Nitrogen adsorption-desorption isotherms were performed at liquid nitrogen
130
temperature (77 K) to determine the pore structure using a gas adsorption analyzer
131
(BEL BELSORP). The specific surface areas were calculated using the Brunauer-
132
Emmett-Teller (BET) equation, and non-local density functional theory (NLDFT) was
133
used to obtain the pore size distributions and pore structure parameters.
134
2.5 Preparation of the electrodes and electrochemical measurements
TE D
M AN U
SC
RI PT
125
Cyclic voltammetry (CV) and galvanostatic charge-discharge (GCD) tests were
136
performed on an IviumStat electrochemical workstation (Ivium Technologies). All the
137
tests were conducted in a three-electrode system with 1.0 M Na2SO4 aqueous solution
138
as the electrolyte; the PPCF and APCFs coated electrode were used as the working
139
electrode, a Ag/AgCl electrode served as the reference electrode, and a platinum coil
140
was used as the counter electrode. The working electrode was prepared by loading a
141
slurry consisting of 80 wt% active materials, 10 wt% carbon black as a conductive
AC C
EP
135
8
ACCEPTED MANUSCRIPT
material, and 10 wt% PVDF as a binder, mixed with NMP, on nickel foam, and then
143
dried at 60 °C for 12 h. The CV curves were obtained in the potential range of −0.5 to
144
0.5 V vs. Ag/AgCl by varying the scan rate from 10 to 100 mV s-1. The GCD
145
measurements were performed at a step-increasing current density from 0.5 to 5 A g-1
146
with a voltage range varying from −0.5 to 0.5 V vs. Ag/AgCl.
147
3. Results and discussion
148
3.1. Surface and structural properties of the APCFs
M AN U
SC
RI PT
142
Fig. 2 shows the SEM images of PPCFs and APCFs, which present uniform
150
distribution of the fiber diameter (thereby indicating a stable electrospinning process);
151
the PCFs have diameters of approximately 300–500 nm [20, 21]. After activation at a
152
low temperature, the average fiber diameters and surface roughness values are nearly
153
the same as that for PPCF up to an activation temperature of 900 °C; however, above
154
1000 °C, thornbush-like carbon fibers are observed in the SEM images (Fig. 2(e) and
155
(f)). Detailed images of the other samples are shown in Fig. S1. The TEM images of the
156
PCFs and as-prepared APCF-1000 (Fig. 3) show similar morphologies as the SEM
157
images. There is an obvious morphological change in the surface of the PCFs before
158
and after KOH activation. Fig. 3(a) and (b) show that the surfaces of the PPCF fibers
AC C
EP
TE D
149
9
ACCEPTED MANUSCRIPT
appear smooth before activation. However, after activation, the surface of APCF-1000
160
(Fig. 3(c) and (d)) became rough, and a marked number of flaws were observed.
161
Furthermore, Fig. 3 (d) shows that the layers of the fibers are transparent at the edge and
162
exhibit a distinct layer separation. Such changes are potentially due to KOH activation,
163
i.e., KOH activates the carbon material mainly by etching and intercalation reactions
164
[22,23].
M AN U
SC
RI PT
159
6KOH + 2C → 2K + 3H2 + 2K2CO3
(1)
166
K2CO3 → K2O + CO2
(2)
167
CO2 + C → 2CO
(3)
168
K2CO3 + 2C → 2K + 3CO
(4)
169
C + K2O → 2K + CO
EP
TE D
165
AC C
(5)
170
The formation of CO2 during the etching reactions also results in the physical activation
171
of the PCFs at a high activation temperature and consequently, contributes to further
172
development of the porosity and specific surface area of the material. Furthermore, the
173
as-prepared metallic K (equations (1), (4), and (5)) can be intercalated into the carbon
174
lattices of the carbon fibers during activation [24] The resulting expansion of the carbon 10
ACCEPTED MANUSCRIPT
175
lattices transforms the morphology of the fibers into a thornbush-like form.
The N2 adsorption-desorption isotherms of the APCFs are presented in Fig. 4.
177
The adsorbed N2 volumes on the PPCFs were very low, indicating their nonporous
178
characteristics. At increased activation temperatures, the resultant APCFs exhibited a
179
drastic uptake of nitrogen at a low relative pressure, which is indicative of the presence
180
of abundant micropores [25-27]. The shapes of the isotherms change from type I to type
181
IV with an increase in the activation temperature. This phenomenon was particularly
182
visible in the isotherms of APCF-900 and APCF-1000, wherein the isotherm shapes
183
indicate an intermediate form between type I and type IV [28]. The isotherms have a
184
steep knee at a low relative pressure and a hysteresis loop at a high relative pressure,
185
which implies that the APCFs contain micropores and mesopores. The thornbush-like
186
morphology of APCF-1000 was assumed to contribute to the development of the pore
187
structure. The porosity parameters for all the samples are presented in Table 1, together
188
with the BET surface areas and pore structure information obtained from the NLDFT
189
results. The yield was determined by measuring the amount of APCF remaining after
190
activation, based on the PPCF mass before activation (Table 1). A higher activation
191
temperature provides a higher electric capacity due to improved pore characteristics.
AC C
EP
TE D
M AN U
SC
RI PT
176
11
ACCEPTED MANUSCRIPT
However, the yield rapidly decreases to less than 10% at 1000 °C, and less than 1% at
193
higher temperatures (≥1000 °C).
RI PT
192
More detailed information on the pore size characteristics is presented in Fig. 5.
195
Fig. 5(a) shows the micro pore size distribution of the samples, which is less than 2 nm.
196
The micropore distributions (0–2 nm) were obtained by the adsorption method (Fig.
197
5(c)), whereas the mesopore distributions (2–50 nm) were obtained by the desorption
198
method (Fig.5 (d)). Increasing the activation temperature of the samples increased the
199
specific surface area, pore volume, and mesopore distributions. The specific surface
200
areas increased steadily with respect to the activation temperature. When a mixture of
201
PCFs and KOH was thermally treated above 1000 °C, the PCFs burnt out and
202
disappeared (yields ≤1%). To elucidate the evolution of porosity with the change in the
203
activation temperature, the change of the cumulative pore volume with the pore size is
204
expressed as a percentage in Fig. 5(b)-(d). Fig. 5(b) shows the change in the cumulative
205
pore volume as a function of the total pore volume in the range of 0-50 nm. APCF-800
206
exhibits the steepest slope, whereas APCF-1000 shows a relatively gentle slope. To
207
obtain detailed information, the change in the cumulative micropore volume in the
208
range 0-2 nm is shown in Fig. 5(c). APCF-800 has the sharpest gradient at
AC C
EP
TE D
M AN U
SC
194
12
ACCEPTED MANUSCRIPT
approximately 1.0 nm or less, and exhibits a narrow pore distribution over 0.7-1.0 nm,
210
which is consistent with its average pore size of 1.6 nm in the 0-50 nm range. The slope
211
of the graph is, on an average, the steepest for APCF-800, and gradually becomes gentle
212
for APCF-1000, which implies that pores developed over a large area as the activation
213
temperature increased. Fig. 5(d) demonstrates the change in the cumulative mesopore
214
volume in the 2-50 nm range; the slopes of APCF-900 and APCF-1000 in the 2-10 nm
215
region are the largest compared to the other samples This confirms that an activation
216
temperature of 900 °C or higher causes pore development in this range. Thus, as the
217
activation temperature increases, high-volume pores develop over a narrow pore range,
218
and the average pore size becomes larger in the 0-50 nm range. Therefore, when APCFs
219
are used as electrode materials, the region of ion movement is widened during charging
220
and discharging, thereby facilitating the free movement of electrolyte ions and thus
221
improving electrical performance.
SC
M AN U
TE D
EP
AC C
222
RI PT
209
XPS measurements were conducted to further investigate the surface properties
223
of the PCFs. The C 1s spectra (Fig. 6) of PPCF and APCF-600 can be deconvoluted into
224
five distinct peaks: sp3-type C-C bonds (284.6 eV), sp2-type C=C bonds (285.3 eV), C-
225
O bonds (286.4 eV), C=N bonds (287.5 eV), and C=(O)O bonds (289.0 eV) [29]. The 13
ACCEPTED MANUSCRIPT
spectra show that the C(O)O peak intensities increased after activation. KOH activation
227
introduced a large number of oxygen-bearing functional groups on the surface of the
228
PCFs, and induced the formation of carbonyl groups (289.0 eV) and other oxygen-
229
bearing groups (286.4 eV) [30]. This was achieved by etching and intercalating the
230
carbon lattice with KOH, resulting in expansion of the PCF husks. The C1s spectrum of
231
APCF-1000 in Fig. 6(d) shows four peaks;the intensity of the peaks increased with
232
increasing activation temperature. The survey scan in Fig. 6(a) indicates that the
233
intensity of C 1s peaks increases more than that of other peaks. The results in Table 2
234
also show that the unstable oxygen and nitrogen contents are preferentially burnt-off
235
during activation at high temperature [31, 32]. This reduction is especially drastic from
236
800 to 900 °C, and APCF-1000 finally exhibits only a few heteroatoms within its lattice.
237
The very low peak intensity of APCF-600 in Fig. S2 clearly demonstrates this
238
temperature effect.
239
3.2. Electrochemical characterization
SC
M AN U
TE D
EP
AC C
240
RI PT
226
The CV curves of the PCFs were measured within the potential window of −0.5
241
to 0.5 V with a series of scan rates from 10 to 100 m V-1 (Fig. 7(a) and (b)). The plots
242
exhibit nearly rectangular cyclic voltammograms, indicating typical electric double14
ACCEPTED MANUSCRIPT
layer capacitive behaviors and pseudocapacitance originating in the surface nitrogen and
244
oxygen functional groups [33, 34]. Although there was a tendency to form shuttle-
245
shaped cyclic voltammograms at higher scan rates, the large rectangular shape for
246
APCF-1000 suggests better electrolyte penetration in the pores and an elevated transient
247
current compared to the PPCFs [35]. The rectangular area of the CV curves was
248
generally proportional to the specific capacitance of the carbon fiber electrode; samples
249
with the largest areas under the CV curves exhibited the best electrochemical efficiency
250
[36].
M AN U
SC
RI PT
243
To further quantify the specific capacitance of the electrodes, GCD tests of the
252
samples were performed at a current density of 1 A g-1 (Fig. 8(a)). The GCD curve of
253
the APCF electrode is nearly triangular, indicating reversible charge/discharge behavior,
254
which agrees with the performance of the EDLCs. However, a slight deviation from the
255
line at a lower potential was observed, which signifies the occurrence of a pseudo-
256
capacitance effect attributed to the heteroatoms attached to the carbon network. APCF-
257
1000 has the highest average pore diameter and a higher specific surface area than other
258
samples (Table 1); this facilitates easy electrode ion movement because of the large
259
pore entrance, leading to high charging behaviors at a high electric current in the CV
AC C
EP
TE D
251
15
ACCEPTED MANUSCRIPT
curves. The high capacitance of APCF-1000 caused by the large average pore size
261
provides more space for ionic movement between the electrode surfaces and electrolytes
262
[37,38]. The specific capacitance (Cs) of the electrode was calculated using equation (6)
263
[39, 43]:
ܥ௦ =
ூ∆௧
SC
264
RI PT
260
∆
(6)
where I, ∆t, m, and ∆V represent the applied current, discharge time, mass of the active
266
material, and the potential window, respectively. The highest specific capacitance
267
obtained from the charge/discharge curves was 103.01 F g-1 for APCF-1000, whereas
268
the lowest was 5.22 F g-1 for the PPCFs at 1 A g-1. The APCFs can deliver 72% of the
269
capacitance, even at a high current density of 5 A g-1. Fig. 8(b) shows the rate
270
performances of the electrode during GCD tests at different current densities. The
271
specific capacitance decreased as the current density increased because of the electrode
272
resistance; therefore, the electrode resistance can be represented by the capacitance
273
retention rate at a high current [40].
AC C
EP
TE D
M AN U
265
274
A long cycle stability of the active materials is critical to the performance of
275
supercapacitors [41-43]. In order to confirm the long-term performance stability of the
276
as-prepared APCF-1000, GCD tests were performed at a current density of 3 A g-1 up to 16
ACCEPTED MANUSCRIPT
3000 cycles (Fig. 9). Remarkably, the as-prepared carbon materials exhibited a high
278
specific capacitance retention of 94% even after 3000 cycles.
279
4. Conclusions
RI PT
277
In summary, a series of APCF electrodes were prepared from PAN powder as
281
precursors via electrospinning and KOH activation. The activation temperature had a
282
significant effect on both, the pore structure and electrochemical performances of the
283
APCFs. As the activation temperature increased, the volume and areas of the micro- and
284
mesopores increased and the distinction became clearer. The APCFs exhibited a unique
285
thornbush-like morphology after activation at high temperatures. The unique
286
microstructure of the samples enabled them to attain a capacitance of 103.01 F g-1 at a
287
current density of 1 A g-1, with a good rate capability and cycle durability, thus
288
projecting them as an excellent electrode material for supercapacitors. The high
289
efficiency of the samples was attributed to the suitable pore structure with a high
290
micropore volume and proper pore size which facilitated smooth movement of the
291
electrolyte ions.
292
Acknowledgements
293
This research was supported by the Technology Innovation Program funded by the
AC C
EP
TE D
M AN U
SC
280
17
ACCEPTED MANUSCRIPT
ministry of Trade, Industry and Energy (MOTIE, Korea) [(10080293, Development of
295
carbon-based non phenolic electrode materials with 3000 m2 g−1 grade specific surface
296
area for energy storage device] and Traditional Culture Convergence Research Program
297
through the National Research Foundation of Korea(NRF) funded by the Ministry of
298
Science, ICT & Future Planning (2018M3C1B5052283).
AC C
EP
TE D
M AN U
SC
RI PT
294
18
ACCEPTED MANUSCRIPT
299
References
300
[1]
Li H, Yuan D, Tang C, Wang S, Sun J, Li Z, Tang T, Wang F, Gong H, He C. Lignin-derived interconnected hierarchical porous carbon monolith with large
302
areal/volumetric capacitances for supercapacitor. Carbon 2016;100:151-7. [2]
carbon fibers for hydrogen storage. Mater. Res. Bull. 2009;44:1871-8.
304 305
Im JS, Park SJ, Lee YS. Superior prospect of chemically activated electrospun
[3]
Xiao J, Wan L, Yang S, Xiao F, Wang S. Design hierarchical electrodes with
SC
303
RI PT
301
highly conductive NiCo2S4 nanotube arrays grown on carbon fiber paper for high-
307
performance pseudocapacitors. Nano Lett. 2014;14:831-8. [4]
and electrodes. J. Power Sources 2011;196:1-12.
309 310
[5]
Xu B, Duan H, Chu M, Cao G, Yang Y. Facile synthesis of nitrogen-doped porous carbon for supercapacitors. J. Mater. Chem. A 2013;1:4565-70.
311 312
Snook GA, Kao P, Best AS. Conducting-polymer-based supercapacitor devices
[6]
Hwang YH, Lee SM, Kim YJ, Kahng YH, Lee K. A new approach of structural
TE D
308
M AN U
306
313
and chemical modification on graphene electrodes for high-performance
314
supercapacitors. Carbon 2016;100:7-15. [7]
[8]
C 111 (2007) 7527-31.
319
[9]
Wang L, Li X, Ma J, Wu Q, Duan X. Non-activated, N, S-co-doped biochar derived from banana with superior capacitive properties. Sustainable Energy
321
2014;2:39-43.
322 323
Subramanian V, Luo C, Stephan AM, Nahm KS, Thomas S, Wei B. Supercapacitors from activated carbon derived from banana fibers, J. Phys. Chem.
318
320
EP
oxide for supercapacitor application. J. Alloy. Compd. 2016;666:15-22.
316 317
Tang J, Shen J, Li N, Ye M. Facile synthesis of layered MnWO4/reduced graphene
AC C
315
[10]
Zhang L, Jiang Y, Wang L, Zhang C, Liu S. Hierarchical porous carbon
324
nanofibers as binder-free electrode for high-performance supercapacitor.
325
Electrochim. Acta 2016;196:189-96. 19
ACCEPTED MANUSCRIPT
326
[11]
Pan H, Yang J, Wang S, Xiong Z, Cai W, Liu J. Facile fabrication of porous
327
carbon nanofibers by electrospun PAN/dimethyl sulfone for capacitive
328
deionization. J. Mater. Chem. A 2015;3:13827-34.
329
[12]
Lee HM, Kim HG, Kang SJ, Park SJ, An KH, Kim BJ. Effects of pore structures on electrochemical behaviors of polyacrylonitrile (PAN)-based activated carbon
331
nanofibers. J. Ind. Eng. Chem. 2015;21:736-40.
332
[13]
RI PT
330
Lee JH, Shin DW, Nam KB, Gim YH, Ko HS, Seo DK, Lee GH, Kim YH, Kim SW, Oh TS, Yoo JB. Continuous bundles of aligned electrospun PAN nano-fiber
334
using electrostatic spiral collector and converging coil. Polymer 2016;84:52-8.
335
[14]
SC
333
An GH, Koo BR, Ahn HJ. Activated mesoporous carbon nanofibers fabricated using water etching-assisted templating for high-performance electrochemical
337
capacitors. Phys. Chem. Chem. Phys. 2016;18:6587-94.
338
[15]
M AN U
336
Lillo-Ródenas MA, Cazorla-Amorós D, Linares-Solano A. Understanding
339
chemical reactions between carbons and NaOH and KOH: an insight into the
340
chemical activation mechanism. Carbon 2003;41:267-75. [16]
distribution in molecular sieve carbon, J. Chem. Eng. Jpn. 1983;16:470-5.
342 343
Horvath G, Kawazoe K, Method for the calculation of effective pore size
TE D
341
[17]
Barret EP, Joyner LG, Hallenda PP, The determination of pore volume and area distributions in porous substances. I. computations from nitrogen isotherms. J. Am.
345
Chem. Soc. 1951;73:373-380. [18]
Amorphous Microporous Materials. Langmuir 2017;33:11138-45.
347 348
[19]
[20]
Seo MK, Park SJ. Electrochemical characteristics of activated carbon nanofiber electrodes for supercapacitors. Mater. Sci. Eng. B 2009;164:106-11.
351 352
Stoeckli F, Centeno TA, Pore size distribution and capacitance in microporous
carbons. Phys. Chem. Chem. Phys. 2012;14:11589-91.
349 350
Kupgan G, Liyana-Arachchi TP, Colina CM, NLDFT Pore Size Distribution in
AC C
346
EP
344
[21]
Tang K, Li Y, Cao H, Su C, Zhang Z, Zhang Y. Amorphous-crystalline
353
TiO2/carbon nanofibers composite electrode by one-step electrospinning for
354
symmetric supercapacitor. Electrochim. Acta 2016;190:678-88. 20
ACCEPTED MANUSCRIPT
355
[22]
Zhang T, Kim CH, Cheng Y, Ma Y, Zhang H, Liu J. Making a commercial carbon
356
fiber cloth having comparable capacitances to carbon nanotubes and graphene in
357
supercapacitors through a “top-down” approach. Nanoscale 2015;7:3285-91.
Mater. Chem. 2012;22:23710-25.
359 360
[24]
Yoon SH, Lim S, Song Y, Ota Y, Qiao W, Tanaka A, Mochida I. KOH activation of carbon nanofibers. Carbon 2004;42:1723-9.
361 362
Wang J, Kaskel S. KOH activation of carbon-based materials for energy storage. J.
[25]
RI PT
[23]
Kierzek K, Frackowiak E, Lota G, Gryglewicz G, Machnikowski J.
SC
358
Electrochemical capacitors based on highly porous carbons prepared by KOH
364
activation. Electrochim. Acta 2004;49:515-23.
365
[26]
M AN U
363
Im JS, Park SJ, Kim TJ, Kim YH, Lee YS, The study of controlling pore size on
366
electrospun carbon nanofibers for hydrogen adsorption. J. Colloid. Interf. Sci.
367
2008;318:42-9. [27]
porous carbon nanofibers. Int. J. Hydrogen Energy 2008;33:4112-5.
369 370
Kim BJ, Lee YS, Park SJ. A study on the hydrogen storage capacity of Ni-plated
[28]
TE D
368
Hsiao HY, Huang CM, Hsu MY, Chen H. Preparation of high-surface-area PAN-
371
based activated carbon by solution-blowing process for CO2 adsorption. Sep. Purif.
372
Technol. 2011;82:19-27. [29]
Guo J, Wen Z, Wang Q, Jin J, Ma G. A conductive selenized polyacrylonitrile
EP
373
cathode material for re-chargeable lithium batteries with long cycle life. J. Mater.
375
Chem. A 2015;3:19815-21.
376
[30]
Sources 2013;234:285-91.
378
[31]
Sevilla M, Valle‐Vigón P, Fuertes AB. N‐Doped Polypyrrole‐Based Porous Carbons for CO2 Capture. Adv. Funct. Mater. 2011;21:2781-7.
380 381
Yun YS, Im C, Park HH, Hwang I, Tak Y, Jin HJ. Hierarchically porous carbon
nanofibers containing numerous heteroatoms for supercapacitors. J. Power
377
379
AC C
374
[32]
J. Ding, Z. Li, K. Cui, S. Boyer, D. Karpuzov, D. Mitlin, Heteroatom enhanced
382
sodium ion capacity and rate capability in a hydrogel derived carbon give record
383
performance in a hybrid ion capacitor. Nano Energy, 23 (2016) 129-137. 21
ACCEPTED MANUSCRIPT
384
[33]
He Y, Wang L, Jia D. Coal/PAN interconnected carbon nanofibers with excellent
385
energy storage performance and electrical conductivity. Electrochim. Acta
386
2016;194:239-45.
387
[34]
Zhang Y, Liu S, Zheng X, Wang X, Xu Y, Tang H, Kang F, Yang QH, Luo J. Biomass Organs Control the Porosity of Their Pyrolyzed Carbon. Adv. Funct.
389
Mater., 2016;27:1604687.
390
[35]
RI PT
388
Xie Q, Zhou S, Zheng A, Xie C, Yin C, Wu S, Zhang Y, Zhao P. Sandwich-like nitrogen-enriched porous carbon/graphene composites as electrodes for aqueous
392
symmetric supercapacitors with high energy density. Electrochim. Acta,
393
2016;189:22-31. [36]
networks for supercapacitor. Chem. Commun. 2013;49:5016-8.
395 396
You B., Wang L., Yao L., Yang J. Three dimensional N-doped graphene–CNT
M AN U
394
SC
391
[37]
Chmiola J., Yushin G., Dash R., Gogotsi Y. Effect of pore size and surface area of
397
carbide derived carbons on specific capacitance. J. Power Sources, 2006;158: 765-
398
72. [38]
Largeot C., Portet C., Chmiola J., Taberna P.L., Gogotsi Y., Simon P. Relation
TE D
399 400
between the ion size and pore size for an electric double-layer capacitor. J. Am.
401
Chem. Soc. 2008;130:2730-1. [39]
[40]
2016;677:1-11.
406
[41]
Kim K.S., Park S.J. Influence of multi-walled carbon nanotubes on the
electrochemical performance of graphene nanocomposites for supercapacitor
408
electrodes. Electrochim. Acta 2011;56:1629-35.
409 410
Sinprachim T., Phumying S., Maensiri S. Electrochemical energy storage performance of electrospun AgOx-MnOx/CNF composites. J. Alloy. Compd.
405
407
EP
electrochemical energy storage. Energy Environ. Sci. 2013;6:1388-414.
403 404
Xu C., Xu B., Gu Y., Xiong Z., Sun J., Zhao X.S. Graphene-based electrodes for
AC C
402
[42]
Yao L., Yang G., Han P., Tang Z., Yang J. Three-dimensional beehive-like
411
hierarchical porous polyacrylonitrile-based carbons as a high performance
412
supercapacitor electrodes. J. Power Sources 2016;315:209-17. 22
ACCEPTED MANUSCRIPT
Liu X., Du G., Zhu J., Zeng Z., Zhu X. NiO/LaNiO3 film electrode with binder-
TE D
M AN U
SC
RI PT
free for high performance supercapacitor. Appl. Surf. Sci. 2016;384:92-8.
EP
414
[43]
AC C
413
23
ACCEPTED MANUSCRIPT
Figure Captions
415
418 419 420 421
carbonization, and activation.
RI PT
417
Fig. 1 Schematic outlining the PPCF preparation process by electrospinning,
Fig. 2 SEM images of samples prepared: (a) PPCF, (b) APCF-600, (c) APCF-800, (d) APCF-900, and (e) & (f) APCF-1000
Fig. 3 TEM characterization of the PCFs before and after KOH activation: (a) & (b)
SC
416
PPCF and (c) & (d) APCF-1000.
Fig. 4 N2 adsorption-desorption isotherms of the samples prepared.
423
Fig. 5 (a) the pore size distribution and (b)-(d) the cumulative pore volume as a
424
M AN U
422
function of the total pore volume of the samples.
Fig. 6 (a) Survey XPS spectra of PPCF, APCF-600, and APCF-1000, and deconvoluted
426
XPS core level peaks of (b) PPCF C 1s, (c) APCF-600 C 1s, and (d) APCF-1000
427
C 1s electrons.
429
Fig. 7 (a) Cyclic voltammograms of the samples at a scan rate of 50 mV s-1 and (b) cyclic voltammogram of APCF-1000.
EP
428
TE D
425
Fig. 8 (a) Galvanostatic charge-discharge curves of the samples at a current density of 1
431
A g-1 and (b) specific capacitance of each sample with respect to the current
432 433 434
AC C
430
density.
Fig. 9 Capacitance retention of APCF-1000 obtained by galvanostatic charge-discharge measurements and the curve shape at a current density of 3 A g-1.
24
RI PT
ACCEPTED MANUSCRIPT
b
SBET (m2 g-1)
d
Vmicro (cm3 g-1)
0.012
0.007
14.1
-
0.058
0.320
3.1
63.2
0.012
0.508
1.6
50.1
0.108
0.614
2.4
39.6
0.940
2.8
5.7
0.021
APCF-600
632
0.375
APCF-800
988
0.533
APCF-900
1220
0.737
APCF-1000
1886
TE D
15
1.196
0.227
AC C
EP
SBET: Specific surface area computed using BET equation at a relative pressure range of 0.0008–0.08. Vtotal: Total pore volume is estimated at a relative pressure PP0-1 = 0.99 using NLDFT model. cV meso: Mesopore (2–50 nm) volume determined from the NLDFT model. dV micro: Micropore (0-2 nm) volume determined from the NLDFT model. e D : Average pore diameter determined by the NLDFT model in range of 0-50 nm. p b
e
VMeso (cm3 g-1)
PPCF
a
c
VTotal (cm3 g-1)
M AN U
a
Samples
SC
Table 1. Textural properties of the samples studied.
Dp (nm)
Yield (%)
SC
Table 2. Elemental properties of the samples studied.
RI PT
ACCEPTED MANUSCRIPT
X-ray photoelectron spectroscopy (at. %)
M AN U
Samples
PPCF APCF-600
APCF-900
O
N
86.79
6.76
6.23
75.75
21.5
2.75
71.54
26.23
1.53
91.86
6.8
1.35
93.13
5.06
1.51
TE D
APCF-800
C
AC C
EP
APCF-1000
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
Fig. 1. Schematic explanation of the PPCF preparation process by electrospinning, carbonization, and activation.
ACCEPTED MANUSCRIPT
(b)
SC
RI PT
(a)
5 µm
M AN U
5 µm
(d)
TE D
(c)
5 µm
AC C
EP
5 µm
Fig. 2. SEM images of samples prepared: (a) PPCF, (b) APCF-600, (c) APCF-800, (d) APCF-900, and (e) & (f) APCF-1000.
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
Fig. 3. TEM characterization of the PCFs before and after KOH activation: (a) & (b) PPCF and (c) & (d) APCF-1000.
SC
APCF-900 APCF-1000
M AN U
600
PPCF APCF-600 APCF-800
500 400 300 200 100 0 0.0
TE D
3
N2 adsorbed volume (cm g
-1
)
700
RI PT
ACCEPTED MANUSCRIPT
0.2
0.4
0.6
0.8 -1
EP
Relative pressure (PP0
)
AC C
Fig. 4. N2 adsorption-desorption isotherms of the samples prepared.
1.0
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Fig. 5. (a) the pore size distribution and (b)-(d) the cumulative pore volume as a function of the total pore volume of the samples.
(b) C 1s
O 1s N 1s
APCF-1000
PPCF
1200
1000
800
600
400
200
Binding energy (eV)
282
284
286
288
C=N C(O)O
290
292
Binding energy (eV)
TE D
C-C
AC C
EP
C=C C-O C=N C(O)O
280
280
0
C-O
(d) C 1s
Intensity (a.u.)
(c) C 1s
C=C
M AN U
APCF-600
C-C
SC
Intensity (a.u.)
C 1s
Intensity (a.u.)
Intensity (a.u.)
(a)
RI PT
ACCEPTED MANUSCRIPT
282
284
286
288
290
Binding energy (eV)
292
280
C-C C=C C-O
282
284
286
288
C(O)O
290
292
Binding energy (eV)
Fig. 6. (a) Survey XPS spectra of PPCF, APCF-600, and APCF-1000, and deconvoluted XPS core level peaks of (b) PPCF C 1s, (c) APCF-600 C 1s, and (d) APCF-1000 C 1s electrons.
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Fig. 7. (a) Cyclic voltammograms of the samples at a scan rate of 50 mV
AC C
EP
TE D
s-1 and (b) cyclic voltammogram of APCF-1000.
ACCEPTED MANUSCRIPT
(a) 0.6
PPCF APCF-600 APCF-800 APCF-900 APCF-1000
RI PT
Voltage (V)
0.4 0.2 0.0
SC
-0.2
0
M AN U
-0.4 50
100
150
200
Time (s)
(b)
PPCF APCF-600 APCF-800
120
TE D
-1
Capacitance (F g )
100
APCF-900 APCF-1000
80
EP
60 40
AC C
20 0
0
1
2
3
4
5
-1
Current density (A g )
Fig. 8. (a) Galvanostatic charge-discharge curves of the samples at a
current of 1 A g-1 and (b) specific capacitance of each sample with respect to the current density.
RI PT SC
100 0.6
80
60
40
M AN U
0.4
Voltage (V)
0.2 0.0 -0.2 -0.4
20
-0.6
TE D
Retention of capacity (%)
ACCEPTED MANUSCRIPT
0
0
500
EP
0
50
100
1000
150
200
250
300
350
Time (s)
1500
2000
2500
3000
Cycle number
Fig. 9. Capacitance retention of APCF-1000 obtained by galvanostatic charge-discharge measurements and
AC C
the curve shape at a current of 3 A g-1.
S1359-8368(18)32525-3
DOI:
10.1016/j.compositesb.2018.10.026
Reference:
JCOMB 6104
To appear in:
Composites Part B
Received Date: 9 August 2018 Revised Date:
7 October 2018
Accepted Date: 9 October 2018
Please cite this article as: Heo Y-J, Lee HI, Lee JW, Park M, Rhee KY, Park S-J, Optimization of the pore structure of PAN-based carbon fibers for enhanced supercapacitor performances via electrospinning, Composites Part B (2018), doi: https://doi.org/10.1016/j.compositesb.2018.10.026. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
1
Manuscript submitted to “Composites Part B: Engineering” as an original paper
2
RI PT
3
4
Optimization of the pore structure of PAN-based carbon fibers for
6
enhanced supercapacitor performances via electrospinning
M AN U
SC
5
7
Young-Jung Heoa, Hyo In Leea, Ji Won Lee, Mira Parkb, Kyong Yop Rheec,** and Soo-
9
Jin Parka,*
a
10
b
12
17
EP
c
Department of Mechanical Engineering, College of Engineering, Kyung Hee University,
AC C
16
Department of Bioenvironmental Chemistry, College of Agriculture & Life Science, Chonbuk National University, Jeonju 54896, South Korea
13
15
Department of Chemistry, Inha University, 100 Inharo, Incheon 22212, Republic of Korea
11
14
TE D
8
Yongin, 17104, Republic of Korea
18
*,** Corresponding authors.
19
Tel.: +82-32-876-7234; Fax: +82-32-867-5604.
20
E-mail addresses : [email protected] (K.Y. Rhee), [email protected] (S.-J. Park).
21 1
ACCEPTED MANUSCRIPT
22
Abstract
Activated microporous polyacrylonitrile-based carbon nanofibers (APCFs)
24
were synthesized by a sequential process of electrospinning, carbonization, and KOH
25
activation. The porosity and surface chemistry of the APCFs strongly depended on the
26
activation temperature. The specific surface area and pore volume varied from 15 to
27
1886 m2 g-1 and 0.021 to 1.196 cm3 g-1, respectively, as the activation temperature
28
increased; this was accompanied by morphology changes at high temperature. The
29
dominant microstructure and minor mesostructure improved the capacitance of carbon.
30
Compared to the other samples, APCFs activated at an optimum temperature of 1000 °C
31
showed the highest specific capacitance of 103.01 F g-1 at 1 A g-1 in 1 mol L-1 Na2SO4
32
aqueous electrolyte, and an excellent cycling durability up to 3000 cycles. The
33
improved electrochemical efficiency could be explained by the high specific surface
34
area, suitable pore size, and influence of heteroatoms relative to the increased electrical
35
double-layers. The change in the pore size distribution with activation temperature is
36
also discussed in detail.
37
AC C
EP
TE D
M AN U
SC
RI PT
23
Keywords: Supercapacitor, Polyacrylonitrile, Porous carbon fibers, KOH activation
38 2
ACCEPTED MANUSCRIPT
39
1. Introduction
Fossil fuels are fast depleting owing to the rapidly growing energy demands,
41
and much study has been dedicated to developing high-efficiency, alternative, and
42
environment friendly energy storage devices such as supercapacitors, fuel cells, and
43
lithium-ion batteries [1-3].
SC
RI PT
40
Supercapacitors, also known as ultracapacitors, have emerged as the most
45
promising energy storage devices for portable electronics and hybrid electric vehicles
46
because of their high power density, long cycle life, and fast charge-discharge rates [4,
47
5]. They have additional advantages such as low maintenance costs and safe operation
48
[6]. Based on the charge-discharge mechanisms, supercapacitors can generally be
49
divided
50
pseudocapacitors. Charge-discharge of EDLCs occurs at the electrode/electrolyte
51
interface by reversible adsorption of electrolyte ions (non-faradaic reactions and
52
physical adsorption). Pseudocapacitors store energy by reversible faradaic redox
53
reactions (chemical) between the active materials and the electrolyte [7, 8].
54
Pseudocapacitors exhibit high capacitance (nearly a dozen times higher than that of
55
EDLCs); however, they are not suitable for industrial applications due to their poor
two
types:
electrical
double-layer
capacitors
(EDLCs)
and
AC C
EP
into
TE D
M AN U
44
3
ACCEPTED MANUSCRIPT
cycle life and expensive active materials used for the production (usually metal oxides
57
such as RuO2, IrO2, and MnO2) [9]. Thus, research has focused on developing
58
commercial supercapacitors based on carbon electrode materials because of their low
59
cost and high stability, despite their low energy density [10]. Various kinds of carbon
60
materials, such as carbon nanofibers, graphene oxide, carbon nanotubes, and activated
61
carbon, have been used as electrode materials [11]. Carbon nanofibers are employed as
62
supercapacitors as they possess excellent chemical stability, good cyclability, and high
63
conductivity.
64
Currently, many carbon materials are produced from polymer precursors, e.g., cellulose-
65
based fibers, polyacrylonitrile (PAN)-based fibers, and pitch-based fibers [12]. PAN, as
66
a precursor of carbon fibers, exhibits a high carbon yield after carbonization and
67
possesses superior mechanical properties [13]. There are several PAN-based fiber
68
preparation techniques, of which the electrospinning technique, which uses electrostatic
69
forces to prepare the fibers, is the most widely used [11, 13]. Electrospinning is a simple
70
and useful method when there are multiple components in the precursor solution, and
71
can produce uniform size-ordered fibers in the nano- to micrometer diameter range.
72
PAN is characterized by good spinnability, which makes it generally suitable as an
AC C
EP
TE D
M AN U
SC
RI PT
56
4
ACCEPTED MANUSCRIPT
electrospinning material. However, pristine PAN-based fibers present poor capacitance
74
owing to their low specific surface area. The specific surface area and porosity are the
75
most important factors for enhanced capacitance, because energy storage in a carbon
76
electrode is grounded by the electrostatic adsorption of electrolyte ions on the large
77
interface between the electrode and the electrolyte [14]. Chemical activation is usually
78
used to increase the specific surface area, for which KOH is the most powerful and
79
accessible activation agent [15]. In pore size analysis of carbon materials, Horvath-
80
Kawazoe (HK) and Barret-Joyner-Hallenda (BJH) methods have been widely used for
81
micropore and mesopore analysis on porous carbons, respectively [16,17]. However,
82
their methods have limitations due to their limited pore range and poor accuracy. In the
83
case of the Density Functional Theory (DFT) method, it has the advantage of being able
84
to analyze from micro to meso size simultaneously with more accurate analysis result
85
than previous methods [18,19].
SC
M AN U
TE D
EP
AC C
86
RI PT
73
In this study, PAN-based carbon fibers (PCFs) were synthesized from PAN
87
powder via electrospinning, and activated at various temperatures with KOH, to prepare
88
activated PAN-based carbon fibers (APCFs). The obtained highly microporous APCFs
89
were used as electrodes in EDLCs, and their electrochemical performance was 5
ACCEPTED MANUSCRIPT
evaluated to optimize the porous structure and thus obtain good electrochemical
91
efficiency.
92
2. Experimental
93
2.1. Materials
SC
RI PT
90
For the electrospinning solution, N,N-dimethylformamide (DMF, ≥99.5%,
95
Daejung Co.) and polyacrylonitrile (PAN, Mw = 150,000, Sigma-Aldrich Co.) were
96
used. KOH (Duksan Pure Chemicals Co.) and HCl (35%, Samchun Pure Chemical Co.)
97
were used as the chemical activation agents. The reagents used to prepare the slurry for
98
the electrochemical measurements were polyvinylidene fluoride (PVDF, Mw = 534,000,
99
Sigma-Aldrich Co.) and 1-methyl-2-pyrrolidone (NMP, ≥99.0%, TCI Co.). All
TE D
M AN U
94
chemicals were used without further purification.
101
2.2. Preparation of PAN-based carbon fibers
AC C
102
EP
100
The PAN solution for electrospinning was prepared by dissolving 10 wt% PAN
103
in DMF by stirring for 12 h at room temperature. The prepared polymer solution was
104
loaded into a 12 mL syringe and ejected from a stainless-steel needle onto an aluminum
105
foil-covered cylindrical collector using an electrospinning apparatus (Nano NC). The 6
ACCEPTED MANUSCRIPT
electrospinning equipment utilized a supplied voltage of 15 kV, feeding rate of 0.8 mL
107
h-1, tip-to-collector distance of 15 cm, and collector speed of 280 rpm, with a syringe
108
pump (KDS-200, KD Scientific). The temperature and relative humidity during
109
spinning process were maintained at 25±2 °C and 50±5%, respectively. The PAN fibers
110
were stabilized at 280 °C for 1 h under air and subsequently carbonized at 1000 °C for 3
111
h under N2.
112
2.3. Chemical activation using potassium hydroxide
M AN U
SC
RI PT
106
The pristine PAN-based carbon fibers (PPCFs) were impregnated with a
114
mixture of distilled water (20 mL), ethanol (5 mL), and KOH (weight ratio of
115
KOH:PPCFs = 4:1), and subsequently dried to a constant weight at 80 °C. The soaked
116
PPCFs were activated at various temperatures (600–1000 °C) for 1 h under N2
117
atmosphere, and labelled as APCF-600, APCF-800, APCF-900, and APCF-1000.
118
Activation temperatures above 1000 °C led to poor yields (≤5%) of APCF. After the
119
activation, the obtained APCFs were first washed with HCl (0.5 M) for neutralization
120
and removal of residual potassium atoms or ions, following which they were filtered
121
with distilled water until the pH of the filtrate reached approximately 6–7. The washed
122
APCFs were dried overnight at 60 °C. A schematic of the PPCF and APCF preparation
123
processes is presented in Fig. 1.
AC C
EP
TE D
113
7
ACCEPTED MANUSCRIPT
124
2.4. Characterization The morphologies of the samples were evaluated using scanning electron
126
microscopy (SEM, Hitachi, SU8010) and transmission electron microscopy (TEM,
127
Philips, CM200). The chemical bond energy on the surfaces and elemental composition
128
were investigated by X-ray photoelectron spectroscopy (XPS, Thermo Scientific, K-
129
Alpha). Nitrogen adsorption-desorption isotherms were performed at liquid nitrogen
130
temperature (77 K) to determine the pore structure using a gas adsorption analyzer
131
(BEL BELSORP). The specific surface areas were calculated using the Brunauer-
132
Emmett-Teller (BET) equation, and non-local density functional theory (NLDFT) was
133
used to obtain the pore size distributions and pore structure parameters.
134
2.5 Preparation of the electrodes and electrochemical measurements
TE D
M AN U
SC
RI PT
125
Cyclic voltammetry (CV) and galvanostatic charge-discharge (GCD) tests were
136
performed on an IviumStat electrochemical workstation (Ivium Technologies). All the
137
tests were conducted in a three-electrode system with 1.0 M Na2SO4 aqueous solution
138
as the electrolyte; the PPCF and APCFs coated electrode were used as the working
139
electrode, a Ag/AgCl electrode served as the reference electrode, and a platinum coil
140
was used as the counter electrode. The working electrode was prepared by loading a
141
slurry consisting of 80 wt% active materials, 10 wt% carbon black as a conductive
AC C
EP
135
8
ACCEPTED MANUSCRIPT
material, and 10 wt% PVDF as a binder, mixed with NMP, on nickel foam, and then
143
dried at 60 °C for 12 h. The CV curves were obtained in the potential range of −0.5 to
144
0.5 V vs. Ag/AgCl by varying the scan rate from 10 to 100 mV s-1. The GCD
145
measurements were performed at a step-increasing current density from 0.5 to 5 A g-1
146
with a voltage range varying from −0.5 to 0.5 V vs. Ag/AgCl.
147
3. Results and discussion
148
3.1. Surface and structural properties of the APCFs
M AN U
SC
RI PT
142
Fig. 2 shows the SEM images of PPCFs and APCFs, which present uniform
150
distribution of the fiber diameter (thereby indicating a stable electrospinning process);
151
the PCFs have diameters of approximately 300–500 nm [20, 21]. After activation at a
152
low temperature, the average fiber diameters and surface roughness values are nearly
153
the same as that for PPCF up to an activation temperature of 900 °C; however, above
154
1000 °C, thornbush-like carbon fibers are observed in the SEM images (Fig. 2(e) and
155
(f)). Detailed images of the other samples are shown in Fig. S1. The TEM images of the
156
PCFs and as-prepared APCF-1000 (Fig. 3) show similar morphologies as the SEM
157
images. There is an obvious morphological change in the surface of the PCFs before
158
and after KOH activation. Fig. 3(a) and (b) show that the surfaces of the PPCF fibers
AC C
EP
TE D
149
9
ACCEPTED MANUSCRIPT
appear smooth before activation. However, after activation, the surface of APCF-1000
160
(Fig. 3(c) and (d)) became rough, and a marked number of flaws were observed.
161
Furthermore, Fig. 3 (d) shows that the layers of the fibers are transparent at the edge and
162
exhibit a distinct layer separation. Such changes are potentially due to KOH activation,
163
i.e., KOH activates the carbon material mainly by etching and intercalation reactions
164
[22,23].
M AN U
SC
RI PT
159
6KOH + 2C → 2K + 3H2 + 2K2CO3
(1)
166
K2CO3 → K2O + CO2
(2)
167
CO2 + C → 2CO
(3)
168
K2CO3 + 2C → 2K + 3CO
(4)
169
C + K2O → 2K + CO
EP
TE D
165
AC C
(5)
170
The formation of CO2 during the etching reactions also results in the physical activation
171
of the PCFs at a high activation temperature and consequently, contributes to further
172
development of the porosity and specific surface area of the material. Furthermore, the
173
as-prepared metallic K (equations (1), (4), and (5)) can be intercalated into the carbon
174
lattices of the carbon fibers during activation [24] The resulting expansion of the carbon 10
ACCEPTED MANUSCRIPT
175
lattices transforms the morphology of the fibers into a thornbush-like form.
The N2 adsorption-desorption isotherms of the APCFs are presented in Fig. 4.
177
The adsorbed N2 volumes on the PPCFs were very low, indicating their nonporous
178
characteristics. At increased activation temperatures, the resultant APCFs exhibited a
179
drastic uptake of nitrogen at a low relative pressure, which is indicative of the presence
180
of abundant micropores [25-27]. The shapes of the isotherms change from type I to type
181
IV with an increase in the activation temperature. This phenomenon was particularly
182
visible in the isotherms of APCF-900 and APCF-1000, wherein the isotherm shapes
183
indicate an intermediate form between type I and type IV [28]. The isotherms have a
184
steep knee at a low relative pressure and a hysteresis loop at a high relative pressure,
185
which implies that the APCFs contain micropores and mesopores. The thornbush-like
186
morphology of APCF-1000 was assumed to contribute to the development of the pore
187
structure. The porosity parameters for all the samples are presented in Table 1, together
188
with the BET surface areas and pore structure information obtained from the NLDFT
189
results. The yield was determined by measuring the amount of APCF remaining after
190
activation, based on the PPCF mass before activation (Table 1). A higher activation
191
temperature provides a higher electric capacity due to improved pore characteristics.
AC C
EP
TE D
M AN U
SC
RI PT
176
11
ACCEPTED MANUSCRIPT
However, the yield rapidly decreases to less than 10% at 1000 °C, and less than 1% at
193
higher temperatures (≥1000 °C).
RI PT
192
More detailed information on the pore size characteristics is presented in Fig. 5.
195
Fig. 5(a) shows the micro pore size distribution of the samples, which is less than 2 nm.
196
The micropore distributions (0–2 nm) were obtained by the adsorption method (Fig.
197
5(c)), whereas the mesopore distributions (2–50 nm) were obtained by the desorption
198
method (Fig.5 (d)). Increasing the activation temperature of the samples increased the
199
specific surface area, pore volume, and mesopore distributions. The specific surface
200
areas increased steadily with respect to the activation temperature. When a mixture of
201
PCFs and KOH was thermally treated above 1000 °C, the PCFs burnt out and
202
disappeared (yields ≤1%). To elucidate the evolution of porosity with the change in the
203
activation temperature, the change of the cumulative pore volume with the pore size is
204
expressed as a percentage in Fig. 5(b)-(d). Fig. 5(b) shows the change in the cumulative
205
pore volume as a function of the total pore volume in the range of 0-50 nm. APCF-800
206
exhibits the steepest slope, whereas APCF-1000 shows a relatively gentle slope. To
207
obtain detailed information, the change in the cumulative micropore volume in the
208
range 0-2 nm is shown in Fig. 5(c). APCF-800 has the sharpest gradient at
AC C
EP
TE D
M AN U
SC
194
12
ACCEPTED MANUSCRIPT
approximately 1.0 nm or less, and exhibits a narrow pore distribution over 0.7-1.0 nm,
210
which is consistent with its average pore size of 1.6 nm in the 0-50 nm range. The slope
211
of the graph is, on an average, the steepest for APCF-800, and gradually becomes gentle
212
for APCF-1000, which implies that pores developed over a large area as the activation
213
temperature increased. Fig. 5(d) demonstrates the change in the cumulative mesopore
214
volume in the 2-50 nm range; the slopes of APCF-900 and APCF-1000 in the 2-10 nm
215
region are the largest compared to the other samples This confirms that an activation
216
temperature of 900 °C or higher causes pore development in this range. Thus, as the
217
activation temperature increases, high-volume pores develop over a narrow pore range,
218
and the average pore size becomes larger in the 0-50 nm range. Therefore, when APCFs
219
are used as electrode materials, the region of ion movement is widened during charging
220
and discharging, thereby facilitating the free movement of electrolyte ions and thus
221
improving electrical performance.
SC
M AN U
TE D
EP
AC C
222
RI PT
209
XPS measurements were conducted to further investigate the surface properties
223
of the PCFs. The C 1s spectra (Fig. 6) of PPCF and APCF-600 can be deconvoluted into
224
five distinct peaks: sp3-type C-C bonds (284.6 eV), sp2-type C=C bonds (285.3 eV), C-
225
O bonds (286.4 eV), C=N bonds (287.5 eV), and C=(O)O bonds (289.0 eV) [29]. The 13
ACCEPTED MANUSCRIPT
spectra show that the C(O)O peak intensities increased after activation. KOH activation
227
introduced a large number of oxygen-bearing functional groups on the surface of the
228
PCFs, and induced the formation of carbonyl groups (289.0 eV) and other oxygen-
229
bearing groups (286.4 eV) [30]. This was achieved by etching and intercalating the
230
carbon lattice with KOH, resulting in expansion of the PCF husks. The C1s spectrum of
231
APCF-1000 in Fig. 6(d) shows four peaks;the intensity of the peaks increased with
232
increasing activation temperature. The survey scan in Fig. 6(a) indicates that the
233
intensity of C 1s peaks increases more than that of other peaks. The results in Table 2
234
also show that the unstable oxygen and nitrogen contents are preferentially burnt-off
235
during activation at high temperature [31, 32]. This reduction is especially drastic from
236
800 to 900 °C, and APCF-1000 finally exhibits only a few heteroatoms within its lattice.
237
The very low peak intensity of APCF-600 in Fig. S2 clearly demonstrates this
238
temperature effect.
239
3.2. Electrochemical characterization
SC
M AN U
TE D
EP
AC C
240
RI PT
226
The CV curves of the PCFs were measured within the potential window of −0.5
241
to 0.5 V with a series of scan rates from 10 to 100 m V-1 (Fig. 7(a) and (b)). The plots
242
exhibit nearly rectangular cyclic voltammograms, indicating typical electric double14
ACCEPTED MANUSCRIPT
layer capacitive behaviors and pseudocapacitance originating in the surface nitrogen and
244
oxygen functional groups [33, 34]. Although there was a tendency to form shuttle-
245
shaped cyclic voltammograms at higher scan rates, the large rectangular shape for
246
APCF-1000 suggests better electrolyte penetration in the pores and an elevated transient
247
current compared to the PPCFs [35]. The rectangular area of the CV curves was
248
generally proportional to the specific capacitance of the carbon fiber electrode; samples
249
with the largest areas under the CV curves exhibited the best electrochemical efficiency
250
[36].
M AN U
SC
RI PT
243
To further quantify the specific capacitance of the electrodes, GCD tests of the
252
samples were performed at a current density of 1 A g-1 (Fig. 8(a)). The GCD curve of
253
the APCF electrode is nearly triangular, indicating reversible charge/discharge behavior,
254
which agrees with the performance of the EDLCs. However, a slight deviation from the
255
line at a lower potential was observed, which signifies the occurrence of a pseudo-
256
capacitance effect attributed to the heteroatoms attached to the carbon network. APCF-
257
1000 has the highest average pore diameter and a higher specific surface area than other
258
samples (Table 1); this facilitates easy electrode ion movement because of the large
259
pore entrance, leading to high charging behaviors at a high electric current in the CV
AC C
EP
TE D
251
15
ACCEPTED MANUSCRIPT
curves. The high capacitance of APCF-1000 caused by the large average pore size
261
provides more space for ionic movement between the electrode surfaces and electrolytes
262
[37,38]. The specific capacitance (Cs) of the electrode was calculated using equation (6)
263
[39, 43]:
ܥ௦ =
ூ∆௧
SC
264
RI PT
260
∆
(6)
where I, ∆t, m, and ∆V represent the applied current, discharge time, mass of the active
266
material, and the potential window, respectively. The highest specific capacitance
267
obtained from the charge/discharge curves was 103.01 F g-1 for APCF-1000, whereas
268
the lowest was 5.22 F g-1 for the PPCFs at 1 A g-1. The APCFs can deliver 72% of the
269
capacitance, even at a high current density of 5 A g-1. Fig. 8(b) shows the rate
270
performances of the electrode during GCD tests at different current densities. The
271
specific capacitance decreased as the current density increased because of the electrode
272
resistance; therefore, the electrode resistance can be represented by the capacitance
273
retention rate at a high current [40].
AC C
EP
TE D
M AN U
265
274
A long cycle stability of the active materials is critical to the performance of
275
supercapacitors [41-43]. In order to confirm the long-term performance stability of the
276
as-prepared APCF-1000, GCD tests were performed at a current density of 3 A g-1 up to 16
ACCEPTED MANUSCRIPT
3000 cycles (Fig. 9). Remarkably, the as-prepared carbon materials exhibited a high
278
specific capacitance retention of 94% even after 3000 cycles.
279
4. Conclusions
RI PT
277
In summary, a series of APCF electrodes were prepared from PAN powder as
281
precursors via electrospinning and KOH activation. The activation temperature had a
282
significant effect on both, the pore structure and electrochemical performances of the
283
APCFs. As the activation temperature increased, the volume and areas of the micro- and
284
mesopores increased and the distinction became clearer. The APCFs exhibited a unique
285
thornbush-like morphology after activation at high temperatures. The unique
286
microstructure of the samples enabled them to attain a capacitance of 103.01 F g-1 at a
287
current density of 1 A g-1, with a good rate capability and cycle durability, thus
288
projecting them as an excellent electrode material for supercapacitors. The high
289
efficiency of the samples was attributed to the suitable pore structure with a high
290
micropore volume and proper pore size which facilitated smooth movement of the
291
electrolyte ions.
292
Acknowledgements
293
This research was supported by the Technology Innovation Program funded by the
AC C
EP
TE D
M AN U
SC
280
17
ACCEPTED MANUSCRIPT
ministry of Trade, Industry and Energy (MOTIE, Korea) [(10080293, Development of
295
carbon-based non phenolic electrode materials with 3000 m2 g−1 grade specific surface
296
area for energy storage device] and Traditional Culture Convergence Research Program
297
through the National Research Foundation of Korea(NRF) funded by the Ministry of
298
Science, ICT & Future Planning (2018M3C1B5052283).
AC C
EP
TE D
M AN U
SC
RI PT
294
18
ACCEPTED MANUSCRIPT
299
References
300
[1]
Li H, Yuan D, Tang C, Wang S, Sun J, Li Z, Tang T, Wang F, Gong H, He C. Lignin-derived interconnected hierarchical porous carbon monolith with large
302
areal/volumetric capacitances for supercapacitor. Carbon 2016;100:151-7. [2]
carbon fibers for hydrogen storage. Mater. Res. Bull. 2009;44:1871-8.
304 305
Im JS, Park SJ, Lee YS. Superior prospect of chemically activated electrospun
[3]
Xiao J, Wan L, Yang S, Xiao F, Wang S. Design hierarchical electrodes with
SC
303
RI PT
301
highly conductive NiCo2S4 nanotube arrays grown on carbon fiber paper for high-
307
performance pseudocapacitors. Nano Lett. 2014;14:831-8. [4]
and electrodes. J. Power Sources 2011;196:1-12.
309 310
[5]
Xu B, Duan H, Chu M, Cao G, Yang Y. Facile synthesis of nitrogen-doped porous carbon for supercapacitors. J. Mater. Chem. A 2013;1:4565-70.
311 312
Snook GA, Kao P, Best AS. Conducting-polymer-based supercapacitor devices
[6]
Hwang YH, Lee SM, Kim YJ, Kahng YH, Lee K. A new approach of structural
TE D
308
M AN U
306
313
and chemical modification on graphene electrodes for high-performance
314
supercapacitors. Carbon 2016;100:7-15. [7]
[8]
C 111 (2007) 7527-31.
319
[9]
Wang L, Li X, Ma J, Wu Q, Duan X. Non-activated, N, S-co-doped biochar derived from banana with superior capacitive properties. Sustainable Energy
321
2014;2:39-43.
322 323
Subramanian V, Luo C, Stephan AM, Nahm KS, Thomas S, Wei B. Supercapacitors from activated carbon derived from banana fibers, J. Phys. Chem.
318
320
EP
oxide for supercapacitor application. J. Alloy. Compd. 2016;666:15-22.
316 317
Tang J, Shen J, Li N, Ye M. Facile synthesis of layered MnWO4/reduced graphene
AC C
315
[10]
Zhang L, Jiang Y, Wang L, Zhang C, Liu S. Hierarchical porous carbon
324
nanofibers as binder-free electrode for high-performance supercapacitor.
325
Electrochim. Acta 2016;196:189-96. 19
ACCEPTED MANUSCRIPT
326
[11]
Pan H, Yang J, Wang S, Xiong Z, Cai W, Liu J. Facile fabrication of porous
327
carbon nanofibers by electrospun PAN/dimethyl sulfone for capacitive
328
deionization. J. Mater. Chem. A 2015;3:13827-34.
329
[12]
Lee HM, Kim HG, Kang SJ, Park SJ, An KH, Kim BJ. Effects of pore structures on electrochemical behaviors of polyacrylonitrile (PAN)-based activated carbon
331
nanofibers. J. Ind. Eng. Chem. 2015;21:736-40.
332
[13]
RI PT
330
Lee JH, Shin DW, Nam KB, Gim YH, Ko HS, Seo DK, Lee GH, Kim YH, Kim SW, Oh TS, Yoo JB. Continuous bundles of aligned electrospun PAN nano-fiber
334
using electrostatic spiral collector and converging coil. Polymer 2016;84:52-8.
335
[14]
SC
333
An GH, Koo BR, Ahn HJ. Activated mesoporous carbon nanofibers fabricated using water etching-assisted templating for high-performance electrochemical
337
capacitors. Phys. Chem. Chem. Phys. 2016;18:6587-94.
338
[15]
M AN U
336
Lillo-Ródenas MA, Cazorla-Amorós D, Linares-Solano A. Understanding
339
chemical reactions between carbons and NaOH and KOH: an insight into the
340
chemical activation mechanism. Carbon 2003;41:267-75. [16]
distribution in molecular sieve carbon, J. Chem. Eng. Jpn. 1983;16:470-5.
342 343
Horvath G, Kawazoe K, Method for the calculation of effective pore size
TE D
341
[17]
Barret EP, Joyner LG, Hallenda PP, The determination of pore volume and area distributions in porous substances. I. computations from nitrogen isotherms. J. Am.
345
Chem. Soc. 1951;73:373-380. [18]
Amorphous Microporous Materials. Langmuir 2017;33:11138-45.
347 348
[19]
[20]
Seo MK, Park SJ. Electrochemical characteristics of activated carbon nanofiber electrodes for supercapacitors. Mater. Sci. Eng. B 2009;164:106-11.
351 352
Stoeckli F, Centeno TA, Pore size distribution and capacitance in microporous
carbons. Phys. Chem. Chem. Phys. 2012;14:11589-91.
349 350
Kupgan G, Liyana-Arachchi TP, Colina CM, NLDFT Pore Size Distribution in
AC C
346
EP
344
[21]
Tang K, Li Y, Cao H, Su C, Zhang Z, Zhang Y. Amorphous-crystalline
353
TiO2/carbon nanofibers composite electrode by one-step electrospinning for
354
symmetric supercapacitor. Electrochim. Acta 2016;190:678-88. 20
ACCEPTED MANUSCRIPT
355
[22]
Zhang T, Kim CH, Cheng Y, Ma Y, Zhang H, Liu J. Making a commercial carbon
356
fiber cloth having comparable capacitances to carbon nanotubes and graphene in
357
supercapacitors through a “top-down” approach. Nanoscale 2015;7:3285-91.
Mater. Chem. 2012;22:23710-25.
359 360
[24]
Yoon SH, Lim S, Song Y, Ota Y, Qiao W, Tanaka A, Mochida I. KOH activation of carbon nanofibers. Carbon 2004;42:1723-9.
361 362
Wang J, Kaskel S. KOH activation of carbon-based materials for energy storage. J.
[25]
RI PT
[23]
Kierzek K, Frackowiak E, Lota G, Gryglewicz G, Machnikowski J.
SC
358
Electrochemical capacitors based on highly porous carbons prepared by KOH
364
activation. Electrochim. Acta 2004;49:515-23.
365
[26]
M AN U
363
Im JS, Park SJ, Kim TJ, Kim YH, Lee YS, The study of controlling pore size on
366
electrospun carbon nanofibers for hydrogen adsorption. J. Colloid. Interf. Sci.
367
2008;318:42-9. [27]
porous carbon nanofibers. Int. J. Hydrogen Energy 2008;33:4112-5.
369 370
Kim BJ, Lee YS, Park SJ. A study on the hydrogen storage capacity of Ni-plated
[28]
TE D
368
Hsiao HY, Huang CM, Hsu MY, Chen H. Preparation of high-surface-area PAN-
371
based activated carbon by solution-blowing process for CO2 adsorption. Sep. Purif.
372
Technol. 2011;82:19-27. [29]
Guo J, Wen Z, Wang Q, Jin J, Ma G. A conductive selenized polyacrylonitrile
EP
373
cathode material for re-chargeable lithium batteries with long cycle life. J. Mater.
375
Chem. A 2015;3:19815-21.
376
[30]
Sources 2013;234:285-91.
378
[31]
Sevilla M, Valle‐Vigón P, Fuertes AB. N‐Doped Polypyrrole‐Based Porous Carbons for CO2 Capture. Adv. Funct. Mater. 2011;21:2781-7.
380 381
Yun YS, Im C, Park HH, Hwang I, Tak Y, Jin HJ. Hierarchically porous carbon
nanofibers containing numerous heteroatoms for supercapacitors. J. Power
377
379
AC C
374
[32]
J. Ding, Z. Li, K. Cui, S. Boyer, D. Karpuzov, D. Mitlin, Heteroatom enhanced
382
sodium ion capacity and rate capability in a hydrogel derived carbon give record
383
performance in a hybrid ion capacitor. Nano Energy, 23 (2016) 129-137. 21
ACCEPTED MANUSCRIPT
384
[33]
He Y, Wang L, Jia D. Coal/PAN interconnected carbon nanofibers with excellent
385
energy storage performance and electrical conductivity. Electrochim. Acta
386
2016;194:239-45.
387
[34]
Zhang Y, Liu S, Zheng X, Wang X, Xu Y, Tang H, Kang F, Yang QH, Luo J. Biomass Organs Control the Porosity of Their Pyrolyzed Carbon. Adv. Funct.
389
Mater., 2016;27:1604687.
390
[35]
RI PT
388
Xie Q, Zhou S, Zheng A, Xie C, Yin C, Wu S, Zhang Y, Zhao P. Sandwich-like nitrogen-enriched porous carbon/graphene composites as electrodes for aqueous
392
symmetric supercapacitors with high energy density. Electrochim. Acta,
393
2016;189:22-31. [36]
networks for supercapacitor. Chem. Commun. 2013;49:5016-8.
395 396
You B., Wang L., Yao L., Yang J. Three dimensional N-doped graphene–CNT
M AN U
394
SC
391
[37]
Chmiola J., Yushin G., Dash R., Gogotsi Y. Effect of pore size and surface area of
397
carbide derived carbons on specific capacitance. J. Power Sources, 2006;158: 765-
398
72. [38]
Largeot C., Portet C., Chmiola J., Taberna P.L., Gogotsi Y., Simon P. Relation
TE D
399 400
between the ion size and pore size for an electric double-layer capacitor. J. Am.
401
Chem. Soc. 2008;130:2730-1. [39]
[40]
2016;677:1-11.
406
[41]
Kim K.S., Park S.J. Influence of multi-walled carbon nanotubes on the
electrochemical performance of graphene nanocomposites for supercapacitor
408
electrodes. Electrochim. Acta 2011;56:1629-35.
409 410
Sinprachim T., Phumying S., Maensiri S. Electrochemical energy storage performance of electrospun AgOx-MnOx/CNF composites. J. Alloy. Compd.
405
407
EP
electrochemical energy storage. Energy Environ. Sci. 2013;6:1388-414.
403 404
Xu C., Xu B., Gu Y., Xiong Z., Sun J., Zhao X.S. Graphene-based electrodes for
AC C
402
[42]
Yao L., Yang G., Han P., Tang Z., Yang J. Three-dimensional beehive-like
411
hierarchical porous polyacrylonitrile-based carbons as a high performance
412
supercapacitor electrodes. J. Power Sources 2016;315:209-17. 22
ACCEPTED MANUSCRIPT
Liu X., Du G., Zhu J., Zeng Z., Zhu X. NiO/LaNiO3 film electrode with binder-
TE D
M AN U
SC
RI PT
free for high performance supercapacitor. Appl. Surf. Sci. 2016;384:92-8.
EP
414
[43]
AC C
413
23
ACCEPTED MANUSCRIPT
Figure Captions
415
418 419 420 421
carbonization, and activation.
RI PT
417
Fig. 1 Schematic outlining the PPCF preparation process by electrospinning,
Fig. 2 SEM images of samples prepared: (a) PPCF, (b) APCF-600, (c) APCF-800, (d) APCF-900, and (e) & (f) APCF-1000
Fig. 3 TEM characterization of the PCFs before and after KOH activation: (a) & (b)
SC
416
PPCF and (c) & (d) APCF-1000.
Fig. 4 N2 adsorption-desorption isotherms of the samples prepared.
423
Fig. 5 (a) the pore size distribution and (b)-(d) the cumulative pore volume as a
424
M AN U
422
function of the total pore volume of the samples.
Fig. 6 (a) Survey XPS spectra of PPCF, APCF-600, and APCF-1000, and deconvoluted
426
XPS core level peaks of (b) PPCF C 1s, (c) APCF-600 C 1s, and (d) APCF-1000
427
C 1s electrons.
429
Fig. 7 (a) Cyclic voltammograms of the samples at a scan rate of 50 mV s-1 and (b) cyclic voltammogram of APCF-1000.
EP
428
TE D
425
Fig. 8 (a) Galvanostatic charge-discharge curves of the samples at a current density of 1
431
A g-1 and (b) specific capacitance of each sample with respect to the current
432 433 434
AC C
430
density.
Fig. 9 Capacitance retention of APCF-1000 obtained by galvanostatic charge-discharge measurements and the curve shape at a current density of 3 A g-1.
24
RI PT
ACCEPTED MANUSCRIPT
b
SBET (m2 g-1)
d
Vmicro (cm3 g-1)
0.012
0.007
14.1
-
0.058
0.320
3.1
63.2
0.012
0.508
1.6
50.1
0.108
0.614
2.4
39.6
0.940
2.8
5.7
0.021
APCF-600
632
0.375
APCF-800
988
0.533
APCF-900
1220
0.737
APCF-1000
1886
TE D
15
1.196
0.227
AC C
EP
SBET: Specific surface area computed using BET equation at a relative pressure range of 0.0008–0.08. Vtotal: Total pore volume is estimated at a relative pressure PP0-1 = 0.99 using NLDFT model. cV meso: Mesopore (2–50 nm) volume determined from the NLDFT model. dV micro: Micropore (0-2 nm) volume determined from the NLDFT model. e D : Average pore diameter determined by the NLDFT model in range of 0-50 nm. p b
e
VMeso (cm3 g-1)
PPCF
a
c
VTotal (cm3 g-1)
M AN U
a
Samples
SC
Table 1. Textural properties of the samples studied.
Dp (nm)
Yield (%)
SC
Table 2. Elemental properties of the samples studied.
RI PT
ACCEPTED MANUSCRIPT
X-ray photoelectron spectroscopy (at. %)
M AN U
Samples
PPCF APCF-600
APCF-900
O
N
86.79
6.76
6.23
75.75
21.5
2.75
71.54
26.23
1.53
91.86
6.8
1.35
93.13
5.06
1.51
TE D
APCF-800
C
AC C
EP
APCF-1000
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
Fig. 1. Schematic explanation of the PPCF preparation process by electrospinning, carbonization, and activation.
ACCEPTED MANUSCRIPT
(b)
SC
RI PT
(a)
5 µm
M AN U
5 µm
(d)
TE D
(c)
5 µm
AC C
EP
5 µm
Fig. 2. SEM images of samples prepared: (a) PPCF, (b) APCF-600, (c) APCF-800, (d) APCF-900, and (e) & (f) APCF-1000.
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
Fig. 3. TEM characterization of the PCFs before and after KOH activation: (a) & (b) PPCF and (c) & (d) APCF-1000.
SC
APCF-900 APCF-1000
M AN U
600
PPCF APCF-600 APCF-800
500 400 300 200 100 0 0.0
TE D
3
N2 adsorbed volume (cm g
-1
)
700
RI PT
ACCEPTED MANUSCRIPT
0.2
0.4
0.6
0.8 -1
EP
Relative pressure (PP0
)
AC C
Fig. 4. N2 adsorption-desorption isotherms of the samples prepared.
1.0
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Fig. 5. (a) the pore size distribution and (b)-(d) the cumulative pore volume as a function of the total pore volume of the samples.
(b) C 1s
O 1s N 1s
APCF-1000
PPCF
1200
1000
800
600
400
200
Binding energy (eV)
282
284
286
288
C=N C(O)O
290
292
Binding energy (eV)
TE D
C-C
AC C
EP
C=C C-O C=N C(O)O
280
280
0
C-O
(d) C 1s
Intensity (a.u.)
(c) C 1s
C=C
M AN U
APCF-600
C-C
SC
Intensity (a.u.)
C 1s
Intensity (a.u.)
Intensity (a.u.)
(a)
RI PT
ACCEPTED MANUSCRIPT
282
284
286
288
290
Binding energy (eV)
292
280
C-C C=C C-O
282
284
286
288
C(O)O
290
292
Binding energy (eV)
Fig. 6. (a) Survey XPS spectra of PPCF, APCF-600, and APCF-1000, and deconvoluted XPS core level peaks of (b) PPCF C 1s, (c) APCF-600 C 1s, and (d) APCF-1000 C 1s electrons.
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Fig. 7. (a) Cyclic voltammograms of the samples at a scan rate of 50 mV
AC C
EP
TE D
s-1 and (b) cyclic voltammogram of APCF-1000.
ACCEPTED MANUSCRIPT
(a) 0.6
PPCF APCF-600 APCF-800 APCF-900 APCF-1000
RI PT
Voltage (V)
0.4 0.2 0.0
SC
-0.2
0
M AN U
-0.4 50
100
150
200
Time (s)
(b)
PPCF APCF-600 APCF-800
120
TE D
-1
Capacitance (F g )
100
APCF-900 APCF-1000
80
EP
60 40
AC C
20 0
0
1
2
3
4
5
-1
Current density (A g )
Fig. 8. (a) Galvanostatic charge-discharge curves of the samples at a
current of 1 A g-1 and (b) specific capacitance of each sample with respect to the current density.
RI PT SC
100 0.6
80
60
40
M AN U
0.4
Voltage (V)
0.2 0.0 -0.2 -0.4
20
-0.6
TE D
Retention of capacity (%)
ACCEPTED MANUSCRIPT
0
0
500
EP
0
50
100
1000
150
200
250
300
350
Time (s)
1500
2000
2500
3000
Cycle number
Fig. 9. Capacitance retention of APCF-1000 obtained by galvanostatic charge-discharge measurements and
AC C
the curve shape at a current of 3 A g-1.