A low ion-transfer resistance and high volumetric supercapacitor using hydrophilic surface modified carbon electrodes

Article abstract of DOI:10.1039/c4ta00158c

A hydrophilic surface modified carbon electrode shows a good electrolyte affinity with homogeneous dispersibility in water, resulting in low ion-transfer resistance and a uniform and dense electrode to give a high volumetric capacitor. The hydrophilic carbon electrode exhibits a superior capacitance (58 F cm^-3, 99.3 mF cm^-2) and is stable up to 5000 cycles. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4ta00158c

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
PAPER
A low ion-transfer resistance and high volumetric
supercapacitor using hydrophilic surface modified
carbon electrodes†
Cite this: J. Mater. Chem. A, 2014, 2,
6663
a
b
b
a
ab
*
Heejoun Yoo, Misook Min, Sora Bak, Yeoheung Yoon and Hyoyoung Lee
Received 10th January 2014
Accepted 18th February 2014
A hydrophilic surface modified carbon electrode shows a good electrolyte affinity with homogeneous
dispersibility in water, resulting in low ion-transfer resistance and a uniform and dense electrode to give
a high volumetric capacitor. The hydrophilic carbon electrode exhibits a superior capacitance (58 F
cmꢀ3, 99.3 mF cmꢀ2) and is stable up to 5000 cycles.
DOI: 10.1039/c4ta00158c
www.rsc.org/MaterialsA
On the other hand, additional strong demands for the power
sources for portable electronic devices require compactness and
Introduction
integration with other electronic components like solid chip
type devices.9–13 In addition, miniaturizing power sources into
solid-chip design devices can potentially increase the density of
electronic devices with various functions and expand their
applications.9,10 Despite these advances, low energy density
remains the biggest challenge and obstacle for the realization of
carbon-based supercapacitors.14 Practically, the volumetric
specic capacitance (F cmꢀ3) and areal specic capacitance
(F cmꢀ2) of supercapacitors based on carbon materials are very
low due to their low density (g cmꢀ3). Likewise, the thickness of
electrodes on supercapacitors with common 2D stack structures
is strongly restricted to a few micrometres and a very low
loading amount due to their low ion diffusion and penetration
properties into the electrodes, and thus they usually exhibit a
very low volumetric specic capacitance (3–50 mF cmꢀ2 or less)
and restrict realization of supercapacitors.5,11,15 According to a
previous study, the laser-scribed graphene electrode shows
4.04 mF cmꢀ2 areal capacitance and less than 0.5 F cmꢀ3
volumetric capacitance with a H2SO4–PVA solid electrolyte.5 The
Recent increase in the use of various portable and exible
electronic devices has generated a growing demand for
advanced energy storage devices that are lightweight, exible,
highly efficient, and require a short charging time.1 Super-
capacitors have received a great deal of attention as next
generation energy storage systems due to their outstanding
characteristics such as high power density, long cycle life,
safety, fast charge–discharge rate, and high efficiency.2–10
Supercapacitors undergo a simple charge storage and release
mechanism based on an electrostatic interaction between the
electrolyte and the electrode surface called electrical double
layer capacitance (EDLC).2–6 The simplicity of this mechanism
of energy storage allows a number of advantages.
For the supercapacitors based on EDLC, low resistance of the
electrode material and its high accessibility with the electrolyte
are the most decisive factors. To achieve high accessibility and
penetration of ions from the electrolyte into the electrode,
several composite designs have been adopted such as gra-
phene–CNT7 (carbon nanotube) and graphene–CDC8 (carbon
derived carbon). CNTs and CDCs can effectively exfoliate gra-
phene sheets and prevent restacking, resulting in enhanced
electrolyte accessibility.2,7,8 Another interesting attempt to
enhance ion accessibility is using in-plane or vertical structures,
which successfully increase ion accessibility and penetration
between two vertically aligned electrodes, resulting in high
capacitance and low ion-transfer resistance compared to the
common 2D-stacked normal supercapacitors.9
all-solid-state micro-SC based on graphene and H3PO4–PVA
shows 0.462 mF cmꢀ2
.
The micro-SC using a graphene and
13
CNT composite electrode exhibits 6.1 mF cmꢀ2 at 10 mV sꢀ1
.
23
Actually, there are many studies concerning the RGO–CNT
composite with superior capacitance performance. Most of
them checked capacitance properties under solution electro-
lytes and 3 electro-measurement systems. Still, the study of
solid-state supercapacitors is rare and shows a very low volu-
metric capacitance even though RGO–CNT composite elec-
trodes are used (see the comparison tables in Table S1† for
RGO–CNT supercapacitors and Table S2† for solid-state
supercapacitors).
Here, we demonstrate a low resistance and high volumetric
on-chip type solid-state supercapacitor (SSC) using hydrophilic
surface modied multi-walled CNTs (MWCNTs) and reduced
graphene oxide (RGO). The hydrophilic surface modied
aNCRI, Center for Smart Molecular Memory, Department of Energy Science,
Sungkyunkwan University, 2066 Seoburo, Jangan-gu, Suwon, Gyeonggi-do 440-746,
Republic of Korea. E-mail: hyoyoung@skku.edu
bNCRI, Center for Smart Molecular Memory, Department of Chemistry, Sungkyunkwan
University, 2066 Seoburo, Jangan-gu, Suwon, Gyeonggi-do 440-746, Republic of Korea
† Electronic supplementary information (ESI) available: Experimental section,
supporting gures, data tables, and movie clips. See DOI: 10.1039/c4ta00158c
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 6663–6668 | 6663

View Article Online
Journal of Materials Chemistry A
Paper
carbon electrode is expected to show good electrolyte affinity Whatman). Aer drying in a vacuum oven (60 ꢁC), the GO–SCNT
and homogeneous dispersibility in water, resulting in low composite lms were reduced by HI vapor reduction with 0.2 ml
resistance and a uniform and dense electrode to give high of HI (57 wt%, Aldrich) and 0.5 ml of acetic acid (99.5 wt%, OCI
volumetric capacity. Especially, for the good accessibility of the Company Ltd) in a 300 ml chamber at 40 ꢁC for 24 h.21 For spray-
electrolyte into the carbon-based electrode, we carefully design RGO–SCNT electrodes, the dispersed mixture solution of GO
the surface modied MWCNTs with hydrophilic sulfonate and SCNTs was sprayed using a spray-gun with a speciꢁc
groups (R–SO3ꢀ), which leads to a hydrophilic surface of the pattern mask on a PET substrate, which was heated to 100 C
outer carbon nanotube, while maintaining high electric during the spraying process. Aer spray coating, the pattern
ꢁ
conductivity through inner carbon nanotubes. Different from electrode was dried under vacuum at 60 C for 24 h. The dried
previous results, this hydrophilic carbon composite lm elec- electrode was then reduced by hydrazine (35%, 0.3 ml, Aldrich)
ꢁ
trode exhibits a much lower series resistance and simulta- vapour reduction in a 300 ml chamber at 90 C for 24 h. Aer
neously lower charge-transfer resistance despite its highly reduction, the electrode was repeatedly washed with D.I. water,
dense structure (over 0.9 g cmꢀ3).
and then dried under vacuum at 60 C for 24 h.
ꢁ
Fabrication of supercapacitor cells
Experimental
To prepare lm electrode supercapacitors, the as-prepared
electrode lms were cut into 10 mm ꢂ 5 mm square pieces. Two
pieces of electrode lms were placed on opposite sides of the
PET substrate with a 0.2–0.3 mm gap and attached with ketone
tape. Pt foil (thickness; 0.1 mm) was used as a current collector,
and was also attached using ketone tape covering the electrode
lms with a width of 0.2 mm. A PVA (10 wt% with H2O) solution
containing 10 wt% of H2SO4 was then cast on the electrodes.
Aer drying on a clean bench for 1 day to remove excess H2O,
electrochemical measurements were carried out. Spray pattern
electrodes were fabricated by the same procedure using Pt foil
and an H2SO4/PVA electrolyte.
Preparation of diazonium salts
4-Benzenediazonium sulfonate was synthesized by diazotiza-
tion of p-sulfonic acid (Aldrich). In a three-necked round
bottom ask, 12.99 g (0.075 mol) of p-sulfonic acid was
dispersed in 1 M HCl. To the well-stirred suspension in an ice-
ꢁ
water bath (3–5 C), a 10% excess of 1 M aqueous solution of
NaNO2 (82.5 ml) was added. Solid p-sulfanilic acid was slowly
dissolved during the addition of NaNO2 and a clear solution was
obtained aer the NaNO2 solution was added completely. The
resulting mixture was then stirred for another 45 min at the
same temperature. A white precipitate was formed and was
collected by ltration, washed with cold water, and then dried
under vacuum.
Characterization
Microstructural characterization was performed using a JEOL
JSM-7404F eld emission scanning electron microscope oper-
ating at 15 kV. XPS measurements were conducted using a
Thermo VG Microtech ESCA 2000 with a monochromatic Al-Ka
X-ray source at 100 W. Water contact angle FT-IR spectra were
collected using a Thermo Nicolet AVATAR 320 instrument.
Preparation of SCNTs
200 mg of MWCNTs were added to a three-necked round
bottom ask with 10 ml of distilled water and 10 ml of ethanol.
The mixture was then sonicated for 30 min and cooled to 0–5 ꢁC
in an ice bath. Next, 10 ml of hypophosphorous acid (H3PO2,
50%, Aldrich) and the prepared 4-benzenediazonium sulfonate
powder (1.2 g) were added to the MWCNT mixture under stir-
ring. Aer stirring for 1 h, an additional 10 ml of H3PO2 aqueous
solution was added. The mixture was then stirred at 5 ꢁC for
another 6 h. The resulting product was ltered with a nylon
membrane (0.45 mm, Whatman) and repeatedly washed with
distilled water. Finally, the resulting powder was dispersed in
D.I. water with 0.5 mg mlꢀ1 by sonication, and le to stand for
2–3 days. The resulting SC solution without sedimentation was
collected and dried at R.T.
Contact angles were measured using
300 microscope.
a SEO Phoenix
Electrochemical characterization
Cyclic voltammograms of SSCs were measured at various scan
rates. The volumetric capacitance derived from CV curves was
Ð
calculated according to the following equation: C ¼ 2( IdV)/
(svDV), where I is the voltammetry current, s is the potential
scan rate, V is the potential for one sweep segment, and v is the
volume of one electrode. The galvanostatic charge–discharge
curves of SSCs were obtained at 2 mA cmꢀ2. The volumetric
capacitance derived from galvanostatic discharge curves was
calculated based on the following equation: C ¼ 2(IDt)/(vDV),
Preparation of lm-RGO–SCNT and spray-RGO–SCNT
electrodes
Graphene oxide (GO) was prepared from natural ake graphite where I is the discharge current, Dt is the time taken for full
(Bay Carbon, SP-1) by an improved method that was previously discharge, v is the volume of one electrode, and DV is the
described.20 GO was dispersed into distilled water at a concen- change in potential during discharge. The effective series
tration of 1 mg mlꢀ1 and a SCNT dispersion was also prepared resistance (ESR) was estimated using the voltage drop at the
at a concentration of 0.5 mg mlꢀ1 in distilled water. The SCNT beginning of the discharge, Vdrop, at a certain constant current
solution was then added to the GO solution under stirring. The Icons, with the formula RESR ¼ Vdrop(2Icons). Energy density (E)
resulting mixture was sonicated for 1 h, and ltered by vacuum and power density (P) were calculated from galvanostatic
ltration with an Anodisc (0.2 mm, with a support ring, discharge curves by the equations E ¼ (1/8)CDV2 and P ¼ E/Dt,
6664 | J. Mater. Chem. A, 2014, 2, 6663–6668
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
where C is the volumetric capacitance, DV is the change in around 1190 cmꢀ1, which corresponded to the symmetrical and
potential during discharge, and Dt is the time taken for full asymmetrical vibrational adsorptions of the SO3 group,
discharge. The electrochemical impedance spectroscopy respectively. The peak observed at 1120 cmꢀ1 is assigned to the
measurements were performed over a frequency range from 106 in-plane bending vibration of the benzene ring.18
to 10ꢀ2 Hz using a CHI660D electrochemical workstation. Based
The XPS spectra from SCNTs exhibited peaks at 168.03 eV
on a RC model, the capacitance was also calculated using the and 532.12 eV, which originated from SO3 groups. The S and O
frequency response analysis, by C ¼ ꢀ1/(2pfZ00), where f is the contents of SCNTs are 2.8 and 13.8 at.%, respectively. Hydro-
frequency in Hz and Z00 is the imaginary part of the impedance, philic-surface treated SCNTs are completely different from
to show the trend of changes in capacitance with frequency.4
normal CNTs, in that they are freely dispersed in water through
simple sonication (Fig. S3a†). The SCNT solution (1 mg mlꢀ1
and 0.01 mg mlꢀ1 in H2O) can be uniformly mixed with a GO
solution, and homogeneous and exible RGO–SCNT lms are
subsequently prepared.
Results and discussion
As expected, the sulfonate functionalized MWCNTs (SCNTs,
Fig. 1a) are well-dispersed in water and freely mixed with gra-
phene oxide (GO) to give a uniform GO–SCNT lm, followed by
reduction of GO–SCNT to nally give reduced graphene oxide
(RGO–SCNT) lms (Fig. 1b). Flexible and uniform lm elec-
trodes (Fig. 1c) are fabricated by vacuum ltration using an
Anodisc, and the specic patterned electrodes on the PET
substrate (Fig. 1d) are prepared through a simple spray coating
process with a patterned mask. The density of RGO–SCNT is
approximately 40 vol% less than that of the normal RGO–CNT.
The resulting high density of the RGO–SCNT lm is attributed
to the homogeneously mixed SCNTs without aggregation into
the graphene sheets, which result in a better packing structure
(Fig. S2†). For the hydrophilic-surface modied SCNTs, the aryl
diazonium salts of sulfanilic acid are prepared as previously
Aer HI vapor reduction, reference RGO lms have a bright
grey color and at surface morphology according to eld
emission scanning electron microscopy (FE-SEM) images on
both the up and down sides of the lm (Fig. 2a). An RGO–
SCNT50 composite lm is prepared by mixing 50 wt% SCNTs
and 50 wt% GO, followed by reduction of the mixed lms. The
RGO–SCNT50 lm shows almost the same black color on both
sides of the lm and exhibits uniformly dispersed CNTs with a
bundled microstructure and less aggregation (Fig. 2b). The
black colored down-side morphology of the RGO–CNT50 lms
is attributed to aggregation of CNT bundles while the grey
colored up side is attributed to exposure of RGO sheets, indi-
cating that the CNT bundles are sparsely mixed with RGO sheets
(Fig. 2c). On the other hand, according to the water contact
angle, RGO lms exhibit hydrophobic surface properties typical
of chemically reduced graphene lms.21 As expected, the surface
of the RGO–SCNT50 lms is highly hydrophilic (water contact
angle; 66.7ꢁ and 56.3ꢁ on the up and down sides of lms,
respectively) due to the presence of SO3H functional groups of
SCNTs, and is signicantly different from those of the RGO and
RGO–CNT50 lms. On the other hand, the down side of RGO–
SCNT50 in comparison with the up side is more hydrophilic due
to higher sedimentation of SCNTs at down-side during the
vacuum ltration process. RGO–CNT50 lms exhibit a more
1
described and conrmed by H NMR (Fig. S1†).16 Surface
functionalization of CNTs with aryl diazonium salts is per-
formed by a chemical diazotization reaction using hypophos-
phorous acid (H3PO2, 50%, Aldrich)17 and the CNTs are
repeatedly washed with a large amount of D.I. water aer the
reaction. The as-prepared SCNTs are then characterized by
Fourier transform infrared spectroscopy (FT-IR) and X-ray
photoelectron spectroscopy (XPS) and compared with CNTs.
FT-IR spectra of SCNTs (Fig. S3b†) reveal differential adsorption
features such as sharp peaks at 1036 cmꢀ1 and a broad peak
Fig. 1 Schematic fabrication process of RGO–SCNT electrodes and
images of (a) SCNTs and (b) RGO–SCNT electrodes. Optical images of Fig. 2 Optical images, SEM images (up and down sides), and water
the (c) film-RGO–SCNT electrode and (d) spray-RGO–SCNT elec- contact angles (up and down sides) of (a) RGO electrode films, (b)
trode on the PET substrate.
RGO–SCNT50 electrode films, and (c) RGO–CNT50 electrode films.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 6663–6668 | 6665

View Article Online
Journal of Materials Chemistry A
Paper
hydrophobic surface (82.4ꢁ and 89.8ꢁ on the up and down sides properties and lower ESR than those of RGO–CNT50 and RGO.
of the lm, respectively), which is due to the strong hydrophobic Fig. 3d shows the specic capacitance versus scan rate from 2 to
properties of CNTs. The bottom-side of RGO–CNT50 in 100 mV sꢀ1. The specic capacitance of RGO–SCNT50 is 44.5 F
comparison with up-side is more hydrophobic due to the exis- cmꢀ3 at 2 mV sꢀ1 and decreases to 20 F cmꢀ3 as the scan rate
tence of more CNTs at the bottom side.
increases to 100 mV sꢀ1. Similarly, the capacitances of RGO–
We investigate the capacitance performance through the in- CNT50 and RGO decrease from 21.4 F cmꢀ3 to 8.4 F cmꢀ3 and
planar symmetrical two-electrode system. Prototype solid-state from 8.2 F cmꢀ3 to 1.2 F cmꢀ3, respectively (Fig. 3d). Frequency
supercapacitor cells (Fig. 3a) are fabricated using as-prepared response analysis (FRA) over a frequency range from 106 to 10ꢀ2
RGO–SCNT50 lm electrodes (5 mm ꢂ 10 mm) and Pt foil Hz yielded the Nyquist plots of RGO–SCNT50, RGO–CNT50 and
(5 mm ꢂ 30 mm ꢂ 0.1 mm) as the current collector and then RGO, as shown in Fig. 3e. Plots of RGO–SCNT50 and RGO–
attached on a PET substrate. The as-prepared solid-state CNT50 featured vertical curves, which are different from RGO,
supercapacitors are evaluated by cyclic voltammetry (CV), gal- indicating nearly ideal capacitive behaviour. The typical Nyquist
vanostatic charge–discharge, and Nyquist impedance plots. plot of the supercapacitor can be classied into three regions
Fig. 3b shows the CV curves of RGO, RGO–CNT50, and RGO– following the frequency range. At high frequency, the intersec-
SCNT50 at 20 mV sꢀ1. RGO exhibits a narrow curve shape, tion point on the real axis (Z0) represents the ohmic resistance of
whereas RGO–CNT50 and RGO–SCNT50 exhibit a rectangular the electrolyte and the internal resistance of the electrode,
curve shape. RGO–SCNT50 has the largest specic capacitance which is described as the series resistance (RS). The semi-
(33.2 F cmꢀ3), while RGO–CNT50 and RGO exhibit capacitance circular plot in the high to middle frequency range corresponds
values of 14 F cmꢀ3 and 3.3 F cmꢀ3 at the same scan rate, to the parallel connection of the ion transfer resistance (RIT) and
respectively. The specic capacitance and ESR calculated from the EDLC. The transition from the semicircle to the long tail is
galvanostatic charge–discharge curves (Fig. 3c and S4,† detailed attributed to the ion diffusion to the inside electrode.22 Based
voltage drop) for RGO–SCNT50 are 29.4 F cmꢀ3 and 29.7 U, on the magnied curve in the high-frequency range (Fig. 3e,
while those of RGO–CNT50 are 12.9 F cmꢀ3 and 43.3 U, inset), the RS and RIT are 20.2 U and 19.2 U for RGO–SCNT50,
respectively. The RGO–SCNT50 exhibits superior capacitance 15.5 U and 53 U for RGO–CNT50, and 60 U and over 380 U for
RGO, respectively. The RS of RGO–CNT50 and RGO–SCNT50 are
lower than that of RGO, suggesting that the enhanced electric
conductivity of the electrode is due to the presence of CNTs and
SCNTs. As expected, the RS of RGO–SCNT50 is slightly higher
than that of RGO–CNT50, which can be explained by the
aromatic SP3 defects of SCNTs resulting from outer carbon
surface functionalization. Furthermore, the RIT values of RGO–
SCNT50 and RGO–CNT50 are much lower than that of RGO and
the slope of the linear plot aer semicircle transition is steeper,
suggesting better accessibility and low ion-transfer resistance of
the electrolyte. In contrast, RGO exhibits a non-ideal capacitive
curve and high RIT, suggesting very poor accessibility of the
polymer electrolyte and high resistance. Our results indicate
that the CNTs are able to function as spacers between RGO
sheets, thereby preventing strong restacking of RGO and
enhancing electrolyte accessibility. Furthermore, the RIT of
RGO–SCNT50 is lower and its curve is steeper than that of RGO–
CNT50, since SCNTs have hydrophilic SO3H functional groups
on their outer layer that provides easy access of protons from
the electrolyte. Simultaneously, the inner layers of SCNTs
maintained high electric conductivity, and thus RS is also
effectively decreased. The capacitance from FRA data versus
frequency based on a series RC model is shown in Fig. 3f. The
capacitance of RGO–SCNT50 is the highest among all the lms
tested in the entire frequency range, while RGO exhibits a very
small capacitance even in a low frequency range. These results
conrm that the superior electrochemical performance of RGO–
SCNT50 composite electrodes is due to the enhanced electric
conductivity and high accessibility of the electrolyte.
Fig. 3 (a) Photo image of proto type solid-state supercapacitor cells,
(b) CV curves of RGO, RGO–CNT50 and RGO–SCNT50 at a scaꢀn2rate
To further evaluate RGO–SCNT electrodes as solid-state
of 20 mV sꢀ1, (c) galvano charge–discharge at constant 2 mA cm , (d)
supercapacitors, we test RGO–SCNT electrodes with different
ratios of SCNTs. As the ratio of SCNTs increased, CV curves
(Fig. 4a) became more rectangular. The CV curves of
volumetric capacitance with different scan rates from 2 to 100 mV sꢀ1
,
(e) Nyquist impedance plots, and (f) frequency response versus
capacitance.
6666 | J. Mater. Chem. A, 2014, 2, 6663–6668
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
contact angles (WCAs) of RGO–SCNT lm electrodes (Fig. S5†)
are slightly different between the up and down sides. Speci-
cally, the down side of RGO–SCNT lms is slightly more
hydrophilic (with lower WCA) than that of the up side, which is
possibly due to accumulation of vacuum force during produc-
tion of the lm electrode by the vacuum lter method. In
addition, the WCA decreased as the ratio of SCNTs increased,
which resulted in increased hydrophilicity. These results are
consistent with our observation of a low ion-transfer resistance
as the ratio of SCNTs increased. Fig. 4d shows the capacitances
from FRA data versus frequency of the RGO–SCNT sample
series; RGO–SCNT50 exhibits the highest capacitance among all
the other samples in the entire frequency range.
We evaluate RGO–SCNT50 for electrochemical stability by
CV at 100 mV sꢀ1. RGO–SCNT50 maintains a specic capaci-
tance of approximately 93% of the initial value aer 5000 cycles
(Fig. S5†), indicating good electrochemical stability. Table S1†
summarizes the specications of electrodes such as volumetric
specic capacitance value, water contact angles, and resistance
properties obtained from Nyquist plots.
To demonstrate on-chip solid-state supercapacitors with
specic patterns, we fabricate RGO–SCNT50 capacitor cells on a
PET substrate using a spray coating method with patterned
Fig. 4 The supercapacitor performance of the RGO–SCNT electrodes
with different contents of SCNTs from 25 to 75 wt%; (a) CV curves at a
scan rate of 20 mV sꢀ1, (b) volumetric capacitance with different scan
rates from 2 to 100 mV sꢀ1, (c) Nyquist impedance plots, and (d)
frequency response versus capacitance.
RGO–SCNT25 show a moderate current increase and decreased mask (Fig. 1d and Movie clip S1†). FE-SEM images of spray-
shape, while RGO–SCNT75 exhibits a more rectangular curve RGO–SCNT50 are shown in Fig. S7b† and compared with lm-
shape. However, the CV curve area of RGO–SCNT75 is smaller RGO–SCNT50 up side (Fig. S7a†). Spray-RGO–SCNT50 has a
than those of both RGO–SCNT25 and RGO–SCNT50, which is coarser surface morphology. In addition, the spray-RGO–
attributed to the active surface area of RGO–SCNT75, which SCNT50 has a lower density (0.7 g cmꢀ3) than that of the RGO–
decreased as the amount of CNTs increased. Specic capaci- SCNT50 lm (0.9 g cmꢀ3). The WCA of the spray-RGO–SCNT50
tance versus scan rate is presented in Fig. 4b. RGO–SCNT25 was 64ꢁ, which is similar to that of the spray-RGO–SCNT50. The
shows the highest capacitance value at a low scan rate, but its capacitance performance of the spray-RGO–SCNT50 evaluated
capacitance sharply decreased from 58.4 F cmꢀ3 to 13 F cmꢀ3 as by CV curves at different scan rates and is similar to that of the
the scan rate increased from 2 to 100 mV sꢀ1, ultimately RGO–SCNT50 lm, suggesting that there is no difference in
decreasing with 77.7%. Likewise, the capacitance of RGO– electrochemical properties for spray coating and vacuum lter
SCNT75 decreased from 17.2 F cmꢀ3 to 12.9 F cmꢀ3, eventually processes (Fig. S6c and S6d†).
decreasing by 25%. RGO–SCNT75 has a lower capacitance than
the other RGO–SCNTs; however, it exhibited a superior rate-
capacitance.
Conclusions
RGO–SCNT50 exhibits a rather modest capacitance decay In summary, we demonstrate low ion-transfer resistance and
curve compared with RGO–SCNT25, exhibiting the highest high volumetric on-chip solid-state supercapacitors consisting
capacitance at scan rates greater than 20 mV sꢀ1. Fig. 4c shows of hydrophilic MWCNTs and graphene akes, and an H2SO4/
Nyquist plots of RGO, RGO–SCNT25, RGO–SCNT50 and RGO– PVA electrolyte. The specially designed hydrophilic MWCNTs
SCNT75. All RGO–SCNT samples exhibit vertical and steeper are prepared by chemical attachment on the outer surface of
curves. Based on magnied curves in the high-frequency range MWCNTs by reduction of para-sulfonic phenyl diazonium salt
(Fig. 4c and inset), RGO–SCNT25 is determined to have a rela- with hypophosphorous acid, and are freely dispersed in water
tively larger semi-circle and a high RCT of over 60 U, while RGO– and can be mixed well with GO solution, ultimately leading to a
SCNT75 exhibits a smaller semi-circle and an RCT of 12.3 U. In uniform RGO–SCNT lm through reduction of a GO–RGO lm.
addition, the plot aer semi-circle transition is steeper as the Hydrophilic treated SCNTs on the outer surface of MWCNTs
amount of SCNTs increases, which suggests better accessibility exhibit a high electric conductivity throughout the inner layers
to the electrolyte. Interestingly, the RS of RGO–SCNT75 (24.6 U) of MWCNTs. Therefore, RGO–SCNT composite electrodes
is higher than that of RGO–SCNT25 (20.8 U) and RGO–SCNT50 exhibit a low wetting angle as well as low ion-transfer resistance.
(20.2 U). This result is possibly due to a low packing density In addition, the homogeneously dispersed RGO–SCNT solution
between RGO sheets and CNT bundles. The electric conductivity in water gave a high volumetric capacitance with high packing
of electrodes is strongly affected by the packing density.19 density and it is easily applied for fabrication of specic
Indeed, the bulk-density of RGO–SCNT75 (0.64 g cmꢀ3) is much patterned electrodes by a spray coating technique, allowing
lower than that of RGO–SCNT25 (0.94 g cmꢀ3) and RGO– integration onto one exible chip with specic patterns. RGO–
SCNT50 (0.9 g cmꢀ3). With respect to hydrophilicity, the water SCNT lms exhibit the highest capacity and rate-capacitance
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 6663–6668 | 6667

View Article Online
Journal of Materials Chemistry A
Paper
properties and they are stable, maintaining a capacitance of
93% of the initial value aer 5000 cycles. In addition, the solid-
state spray-RGO–SCNT electrodes can be connected in series or
in parallel to improve the overall output potential and/or
current. In addition, they may also be integrated onto various
substrates for energy storage devices, including portable,
stretchable, and wearable electronic devices.
E. Bekyarova, M. E. Itkis and R. C. Haddon, Adv. Energy
Mater., 2012, 2, 438.
8 M. Heon, S. Lo, J. Applegate, R. Nolte, E. Cortes,
J. D. Hettinger, P.-L. Taberna, P. Simon, P. Huang,
M. Brunet and Y. Gogotsi, Energy Environ. Sci., 2011, 4, 135;
Z. Lei, Z. Liu, H. Wang, X. Sun, L. Lu and X. S. Zhao,
J. Mater. Chem. A, 2013, 1, 2313.
9 J. J. Yoo, K. Balakrishnan, J. Huang, V. Meunier,
B. G. Sumpter, A. Srivastava, M. Conway, A. L. M. Reddy,
J. Yu, R. Vajtai and P. M. Ajayan, Nano Lett., 2011, 11, 1423.
Acknowledgements
This work was supported by the National Research Foundation 10 J. Chmiola, C. Largeot, P. L. Taberna, P. Simon and
of Korea (NRF) grant funded by the Korean government (MSIP)
(Grant no. 2006-0050684 and Grant no. B551179-13-01-01).
Y. Gogotsi, Science, 2010, 328, 480.
11 M. Kaempgen, C. K. Chan, J. Ma, Y. Cui and G. Gruner, Nano
Lett., 2009, 9, 1873.
12 M. Beidaghi and C. Wang, Adv. Funct. Mater., 2012, 22, 4501.
13 Z. Niu, L. Zhang, L. Liu, B. Zhu, H. Dong and X. Chen, Adv.
Mater, 2013, 25, 4035.
14 Y. Xu, Z. Lin, X. Huang, Y. Liu, Y. Huang and X. Duan, ACS
Nano, 2013, 7, 4042.
15 M. D. Stoller and R. S. Ruoff, Energy Environ. Sci., 2010, 3,
1294; Z. Weng, Y. Su, D.-W. Wang, F. Li, J. Du and
H.-M. Cheng, Adv. Energy Mater., 2011, 1, 917.
Notes and references
1 D. Pech, M. Brunet, H. Durou, P. H. Huang, V. Mochalin,
Y. Gogotsi, P. L. Taberna and P. Simon, Nat. Nanotechnol.,
2010, 5, 651; Z. Niu, H. Dong, B. Zhu, J. Li, H. H. Hng,
W. Zhou, X. Chen and S. Xie, Adv. Mater., 2013, 25, 1058;
H. Yoo, Y. Kim, J. Lee, H. Lee, Y. Yoon, G. Kim and H. Lee,
Chem.–Eur. J., 2012, 18, 4923.
2 H. M. Jeong, J. W. Lee, W. H. Shin, Y. J. Choi, H. J. Shin, 16 Q. Li, L. Dong, F. Sun, J. Huang, H. Xie and C. Xiong, Chem.–
J. K. Kang and J. W. Choi, Nano Lett., 2011, 11, 2472;
Eur. J., 2012, 18, 7055.
Y. Zhai, Y. Dou, D. Zhao, P. F. Fulvio, R. T. Mayes and 17 X. Wang, R. Liu, M. M. Waje, Z. Chen, Y. Yan, K. N. Bozhilov
and P. Feng, Chem. Mater., 2007, 19, 2395;
M. Pandurangappa, N. S. Lawrence and R. G. Compton,
Analyst, 2002, 127, 1568.
S. Dai, Adv. Mater., 2011, 23, 4828.
3 P. Simon and Y. Gogotsi, Nat. Mater., 2008, 7, 845;
M. D. Stoller, S. Park, Y. Zhu, J. An and R. S. Ruoff, Nano
Lett., 2008, 8, 3498.
4 Y. Zhu, S. Murali, M. D. Stoller, K. J. Ganesh, W. Cai,
P. J. Ferreira, A. Pirkle, R. M. Wallace, K. A. Cychosz,
18 C. Y. Du, T. S. Zhao and Z. X. Liang, J. Power Sources, 2008,
176, 9; J. C. Yang, M. J. Jablonsky and J. W. Mays, Polymer,
2002, 43, 5125.
M. Thommaes, D. Su, E. A. Stach and R. S. Ruoff, Science, 19 I. K. Moon, J. I. Kim, H. Lee, W. C. Kim and H. Lee, Sci. Rep.,
2011, 332, 1537.
2013, 3, 1112.
20 D. C. Marcano, D. V. Kosynkin, J. M. Berlin, A. Sinitskii,
Z. Sun, A. Slesarev, L. B. Alemany, W. Lu and J. M. Tour,
ACS Nano, 2010, 4, 4806.
21 I. K. Moon, J. Lee, R. S. Ruoff and H. Lee, Nat. Commun.,
2010, 1, 73.
22 B. G. Choi, J. Hong, W. H. Hong, P. T. Hammond and
H. Park, ACS Nano, 2011, 5, 7205; C. Masarapu, H. F. Zeng,
K. H. Hung and B. Wei, ACS Nano, 2009, 3, 2199.
23 M. Beidaghi and C. Wang, Adv. Funct. Mater., 2012, 22,
4501.
5 M. F. El-Kady, V. Strong, S. Dubin and R. B. Kaner, Science,
2012, 335, 1326.
6 Y. Xu and G. Shi, J. Mater. Chem., 2011, 21, 3311; Z.-S. Wu,
A. Winter, L. Chen, Y. Sun, A. Turchanin, X. Feng and
K. Mullen, Adv. Mater., 2012, 24, 5130; Y. Li, Z. Li and
P. K. Shen, Adv. Mater., 2013, 25, 2474.
7 A. L. M. Reddy, S. R. Gowda, M. M. Shaijumon and
P. M. Ajayan, Adv. Mater., 2012, 24, 5045; Z. Fan, J. Yan,
L. Zhi, Q. Zhang, T. Wei, J. Geng, M. Zhang, W. Qian and
F. Wei, Adv. Mater., 2010, 22, 3723; N. Jha, P. Ramesh,
6668 | J. Mater. Chem. A, 2014, 2, 6663–6668
This journal is © The Royal Society of Chemistry 2014