Materials Chemistry A: Journal of

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

Journal of

Materials Chemistry A
View Article Online
PAPER View Journal | View Issue

Carbon–TiO2 composites as high-performance


supercapacitor electrodes: synergistic effect
Published on 27 November 2017. Downloaded by Ege Universitesi on 8/21/2020 11:33:47 AM.

Cite this: J. Mater. Chem. A, 2018, 6,


633 between carbon and metal oxide phases†
A. Elmouwahidi, E. Bailón-Garcı́a, * J. Castelo-Quibén, A. F. Pérez-Cadenas,
F. J. Maldonado-Hódar and F. Carrasco-Marı́n

A series of carbon xerogels doped with different percentages of TiO2 has been studied as a tentative means
of preparing electrodes for supercapacitors. Carbon composites were obtained by an inverse emulsion
method in n-heptane, and after carbonization at 900  C the metal oxide phase was well dispersed in the
carbon phase with a crystal size of less than 4 nm, and only an anatase phase was detected. An increase
in the percentage of TiO2 produced a decrease in the hydrophobicity of the composite, which improved
the wettability of the electrodes. XPS results showed that Ti3+ and Ti4+ were present on the surface of
samples, and the presence of both oxidation states can improve the electron mobility in the inorganic
phase. The obtained composite materials possessed specific surface areas that ranged from 423 to
539 m2 g1 and very well developed micro- and mesoporosity with a total pore volume that ranged from
0.361 to 0.480 cm3 g1. The mean size of supermicropores increased as the percentage of TiO2
increased, whereas practically no variation was found in the size of ultramicropores. Two-electrode
symmetric supercapacitors based on the carbon xerogel–TiO2 composites exhibited high
electrochemical performance, which was better than that of other similar materials in the literature, and
displayed high capacitance (up to 137 F g1 at 0.250 A g1 for the composite containing 20% TiO2), high
capacitance retention (66–80%) at 20 A g1 and a high energy density of 20.15 W h kg1 at a power
Received 12th September 2017
Accepted 27th November 2017
density of 138.11 W kg1 in the voltage range of 0 V to 1.1 V. A sample with a combination of low
hydrophobicity and an adequate micro/mesopore network with an intermediate content of TiO2
DOI: 10.1039/c7ta08023a
exhibited the best performance for energy storage. A floating test showed the very good cyclability of
rsc.li/materials-a the synthesized materials.

different carbon nanostructures.5–7 Currently, the electrodes of


Introduction most commercial supercapacitors are made of carbon because
The current demands for clean and sustainable energy, together it is inexpensive, has good resistance to corrosion and has
with their advantages of a high power density, high efficiency, excellent cycling stability and a long service lifetime, because
and long life expectancy, have made electrochemical super- the electrodes undergo no chemical changes during the charge/
capacitors among the major emerging devices for energy discharge processes.5,8 However, the active surface area of the
storage and power supply.1 Supercapacitors, which are electrode and the pore size distribution limit the maximum
commonly known as electric double-layer capacitors (EDLCs), capacitance.
have a higher energy density than conventional dielectric A composite electrode with a nano-sized metal oxide incor-
capacitors owing to the large surface area of the porous elec- porated in a highly conductive carbon support would be
trode materials. They also have a higher power density and low advantageous for designing supercapacitor electrode materials.
temperature dependence, as well as nearly unlimited cyclability Over the past few years, various metal oxides, such as manga-
(>100 000 cycles).2–4 nese, ruthenium, and zirconium oxides, MnO2, NiO, TiO2,
In the past decade, various carbon–metal oxide composite Fe2O3 and their hydroxides have been used to prepare metal
electrodes have been developed by integrating metal oxides into oxide–carbon composites.5–7,9–11 Titanium dioxide has been
used for many applications owing to its semiconducting prop-
erties, accessible surface, electrochemical behaviour and long-
Research Group in Carbon Materials, Inorganic Chemistry Department, Faculty of term chemical stability, which include the preparation of elec-
Sciences, University of Granada, Campus Fuente Nueva s/n. 18071, Granada, Spain.
trochemical composite electrodes.
E-mail: estherbg@ugr.es
† Electronic supplementary information (ESI) available. See DOI:
TiO2 is an interesting electrode material for electrochemical
10.1039/c7ta08023a energy storage systems because of the ideal capacitive response

This journal is © The Royal Society of Chemistry 2018 J. Mater. Chem. A, 2018, 6, 633–644 | 633
View Article Online

Journal of Materials Chemistry A Paper

of TiO2 (rectangular cyclic voltammetry curves) observed in capacitance obtained from air-annealed TiO2 nanotubes in the
some works,12–14 which indicates that the storage mechanism in same conditions, which can be ascribed to the increased elec-
TiO2 comprises conventional electric double-layer storage. trode conductivity associated with the active hydrogen-
However, owing to the low conductivity of TiO2, the preparation generated Ti3+ sites; however, the capacitance also sharply
of a titanium oxide–carbon composite can be an effective declined when the scan rate was increased from 10 to 1000 mV
approach for increasing the capacitance and rate capability of s1. Therefore, in spite of the increase achieved in the capaci-
TiO2. The main challenge is the relatively low theoretical tance, it is known that the incorporation of Ti3+ sites by this self-
capacity of TiO2, which restricts its further practical applica- doping process (plasma treatment or hydrogenation reactions)
tions because of its wide electronic band gap (3.2 eV), and its will generate abundant disordered states and lead to a negative
high resistivity, which would lead to high internal resistance in effect on cycling stability.15 Therefore, alternative methods of
Published on 27 November 2017. Downloaded by Ege Universitesi on 8/21/2020 11:33:47 AM.

charge storage devices. Consequently, the specic capacitance incorporating Ti3+ sites in TiO2 must be studied in order to
of metal oxide-based supercapacitors declines considerably increase the capacitance while maintaining high cycling
upon an increase in the scan rate owing to their poor electronic stability. In this context, our research group has pointed out
conductivity. Therefore, it is necessary to combine metal oxides that in composites prepared by sol–gel methods owing to inti-
with other materials that have a high conductivity to form mate contact between the carbon and titania phases high
composites. To resolve these issues, persistent efforts have been dispersion of anatase TiO2 nanoparticles is achieved, and these
focused on modifying pure TiO2 by introducing oxygen vacan- nanoparticles are partially reduced with a surface distribution
cies (Ti3+ sites),15–17 doping with heteroatoms,18–21 or incorpo- of oxygen highly dependent on the TiO2 content. This interac-
rating carbon materials or dyes18,22–28 in order to reduce the tion also causes the band gaps of the C–TiO2 composites to be
band gap and efficiently increase the electrical conductivity. lower than 2.8 eV, which improves their conductivity and pho-
Nonetheless, the TiO2-based electrode materials that have been tocatalytic response.32,33
reported still do not meet the requirements of high- Consequently, the combination of TiO2 with carbon mate-
performance energy storage devices.29 Moreover, overcoming rials could be an interesting option for developing stable elec-
problems such as charge recombination and the thermal/ trode materials with increased specic capacitance. Several
electrochemical instability that arises from the incorporation TiO2–carbon composites have been developed that were based
of dopants is still a great challenge.30 mainly on graphene and carbon nanotubes.23–25,28,34–36 However,
In addition, special attention has been paid to the creation of many of the processes that were developed to produce these
oxygen vacancies (Ti3+ sites) in TiO2 by plasma treatment and carbon-based EDLCs involve environmentally unfriendly
hydrogenation reactions. The increase in capacitance can be chemicals and elaborate procedures. Furthermore, the high
attributed to the combined contribution from the increased cost of carbon nanotubes and graphene might render them
carrier density (attributed to the increase in oxygen vacancy unfeasible for practical everyday applications.31
states, which are known to be electron donors for TiO2) and the With this base, this paper describes the preparation of
increased density of surface hydroxyl groups.17 For instance, Wu carbon xerogel–titanium oxide composites by an inverse-
et al.15 demonstrated that titania nanotubes treated by H2 plasma emulsion sol–gel method and the study of the effect of the
illumination (ATO-H) displayed a rough and amorphous layer on titanium percentage on the porosity, structure and electro-
the surface of the nanotubes with the simultaneous incorpora- chemical performance of supercapacitors prepared from these
tion of Ti3+ and OH groups, and consequently, at a current materials. These materials exhibited excellent capacity and
density of 0.05 mA cm2 in charge–discharge measurements, the cycling performance that were much better than those of other
specic capacitance of the ATO-H electrode substantially similar materials in the literature.
increased by a factor of z7.4 in comparison with that of pristine
nanotubes, with a value of as high as 7.22 mF cm2. Similarly, Experimental
Salari et al.31 demonstrated an increase in specic capacitance
by optimizing thermal treatment under an argon atmosphere. Synthesis of TiO2–carbon xerogel composites
The increased specic capacitance could be attributed to TiO2–carbon xerogel composites were prepared by mixing
partially reduced valence states created on the surface of the resorcinol–formaldehyde as the carbon source and titanium(IV)
titania nanotubes. The increased capacitance might also be due isopropoxide as the titanium oxide precursor. In detail, Span 80
to the presence of other defects formed during annealing in (S) was dissolved in 900 mL of n-heptane and heated at 70  C
a reductive atmosphere, such as titanium interstitials. Although under reux and stirring (450 rpm). Then a mixture containing
the specic capacitance (i.e., capacitance per unit planar area) resorcinol (R), formaldehyde (F) and water (W) was added drop
of 2.6 mF cm2 obtained at a scan rate of 1 mV s1 was high in by drop to the above solution. Aer this addition, an amount of
comparison with previous reports for TiO2 (10–120 mF cm2), it titanium isopropoxide was added drop by drop to the mixture.
declined to about 0.6 mF cm2 at a higher scan rate of The molar ratios of the mixture were R/F ¼ 1/2, R/W ¼ 1/14 and
100 mV s1, which indicated poor rate capability. More recently, R/S ¼ 4.5.
a substantial improvement in capacitive performance was The mixture was aged at 70  C for 24 h under stirring before
realized in Ti nanotubes hydrogenated at high temperatures.17 A being ltered, and the solid that was obtained was placed in
specic capacitance of 3.24 mF cm2 at a scan rate of 100 mV acetone (5 days, changing the acetone twice daily) to remove the
s1 was obtained, which was 40 times higher than the Span that was used and exchange water within the pores for

634 | J. Mater. Chem. A, 2018, 6, 633–644 This journal is © The Royal Society of Chemistry 2018
View Article Online

Paper Journal of Materials Chemistry A

acetone. This procedure reduced the collapse of pores during the scanned from 20 to 70 . The chemical characterization of the
subsequent drying process. Then, the gel was ltered again and xerogels was further performed by X-ray photoelectron spectros-
dried by microwave heating under an argon atmosphere for copy (XPS). The spectra were recorded using a Kratos Axis Ultra
periods of 1 minute at 300 W until its weight was constant using DLD X-ray photoelectron spectrometer equipped with a hemi-
a Saivod MS-287 W microwave oven. Pyrolysis of the organic spherical electron analyzer connected to a delay-line detector
xerogel–titanium oxide composite to obtain the corresponding (DLD).
carbon xerogel–titanium oxide composite was carried out at
900  C in a tubular furnace using a N2 ow of 300 cm3 min1 and Electrochemical measurements
a heating rate of 1  C min1 to allow the easy removal of pyrolysis
Electrodes were prepared by combining 90% CTiX and 10%
gases and a dwell time of 2 h at this temperature. Depending on
Published on 27 November 2017. Downloaded by Ege Universitesi on 8/21/2020 11:33:47 AM.

polytetrauoroethylene (PTFE) to form a homogeneous mixture.


the amount of titanium oxide (10, 20, 30 and 40 wt% of the
The mixture was dried overnight at 80  C, and then 5 mg of the
carbonized materials), four carbon xerogel–titanium oxide
dried mixture was pasted onto discs of graphite paper with
composites were prepared by adjusting the ratio of the alkoxide,
a diameter of 5 mm. Two-electrode cells were assembled with
assuming a weight loss during carbonization of 50%. The carbon
two CTiX electrodes sandwiching a separator comprising
xerogel–titanium oxide composites that were prepared were
a porous glassy brous material. In order to determine the
denoted as CTiX (X corresponds to the theoretical percentage of
pseudofaradaic contribution of TiO2, a three-electrode system
titanium oxide present in the carbonized composite, e.g., CTi40
was also used to characterize the samples. For this purpose,
contains 40 wt% of TiO2). Moreover, a pure carbon xerogel (C100)
a typical three-electrode cell with Ag/AgCl as the reference
was prepared by following the same method but without adding
electrode and a Pt wire as the counter electrode was employed.
the titanium precursor for use as a reference material.
Cyclic voltammetry (CV), galvanostatic charge–discharge
(GCD) tests, electrochemical impedance spectroscopy (EIS) and
Textural and chemical characterization charge–discharge cycles for testing long-term stability were
performed on electrodes produced from carbon–titanium
Textural characterization was carried out by the adsorption of composites in 1 M sulfuric acid as the electrolyte using a Bio-
N2 and CO2 at 196  C and 0  C, respectively, using a Quan- logic VMP-300 potentiostat/galvanostat.
tachrome Autosorb analyzer. The BET and Dubinin–Radush- Cyclic voltammetry tests (CV) were carried out within the
kevich equations were used to determine the apparent surface range of 0 to 1.1 V using scan rates of 0.5, 2.5, 5, 10, 20 and
area (SBET) and micropore volume (W0) and the mean micropore 30 mV s1, and the capacitance was calculated using the
width (L0) and micropore surface area (Smic), respectively. following relation (eqn (2)):39,40
Furthermore, the BJH method was used to calculate the meso- X
pore volume of the samples (Vmes). Pore size distributions were jIjDt
C¼ (2)
also determined by employing the BJH method. The total pore mDV
X
volume was considered to be the volume of N2 adsorbed at P/P0 where jIjDt is the area under the curve of current (A) against
¼ 0.95. The morphology of the samples was studied by scanning
time (s), m is the total mass of active materials in both elec-
electron microscopy (SEM) and high-resolution transmission
trodes (g), and DV is the potential window (V).
electron microscopy (TEM) using a LEO (Carl Zeiss) Gemini
Galvanostatic charge–discharge analysis was carried out at
1530 microscope and an FEI Titan G2 microscope, respectively.
different current rates from 125 mA g1 to 20 A g1, and the
The amount of TiO2 in the samples was determined by
gravimetric capacitance was calculated using eqn (3):41,42
thermogravimetric analysis (TGA). TGA was performed in
owing air at a heating rate of 10  C min1 using a Mettler Id Dt
C¼ (3)
Toledo TGA/DSC1 thermogravimetric analyzer. mDV
In order to determine the hydrophobicity of the samples, the where Id is the discharge current, Dt is the discharge time, m is
enthalpies of immersion of the carbon phase into benzene and the total mass of AC in the electrodes, and DV is the voltage
water, i.e., DiH(C6H6) and DiH(H2O), were determined with window aer the ohmic drop is subtracted. The coulombic
a calorimeter of the Tian–Calvet type.37,38 The hydrophobicity of efficiency (%) from chronopotentiometric measurements was
the samples was calculated according to eqn (1). The enthalpy of calculated from the discharge and charge times td and tc,
immersion into water depends upon specic and non-specic respectively, by the formula (td/tc)  100.
interactions; in contrast, in the case of benzene the enthalpy The impedance spectroscopy measurements were carried
of immersion depends only on the surface area of the carbon out in the frequency range from 100 kHz to 0.01 Hz at an AC
phase accessible to benzene molecules. amplitude of 10 mV. The value of the capacitance C from EIS
Di HðC6 H6 Þ  Di HðH2 OÞ measurements was obtained using eqn (4):
HF ¼ (1)
Di HðC6 H6 Þ Z 00
C¼ (4)
2pfZ 2
The crystalline phases of the resulting samples were charac- where f is the frequency and Z2 ¼ Z0 2 + Z00 2, where Z0 and Z00 are
terized using a Bruker D8 Advance X-ray diffractometer with Cu Ka the real and imaginary parts of the complex impedance,
radiation at a wavelength (l) of 1.541 Å. The 2q angles were respectively.

This journal is © The Royal Society of Chemistry 2018 J. Mater. Chem. A, 2018, 6, 633–644 | 635
View Article Online

Journal of Materials Chemistry A Paper

For reasons of comparison, the gravimetric capacitances in


the 2EC determined by the three techniques were multiplied by
4 to obtain the capacitances per single electrode, which were the
equivalents for the 3EC43 and were denoted as CCV, CCP and CEIS,
respectively.
However, by convention, the gravimetric capacitances ob-
tained using eqn (3) for the 2EC were used to calculate the
energy density via a Ragone plot using eqn (5), and the power
density was calculated using eqn (6).44,45
 2
Published on 27 November 2017. Downloaded by Ege Universitesi on 8/21/2020 11:33:47 AM.

Ceq:3 Vmax  IRdrop


E¼ (5)
2
 
  E IðAÞ  Vmax  IRdrop
P W kg1 ¼ ¼ (6)
Dt 2m ðkgÞ

Results and discussion Fig. 1 (a) SEM, (b) SEM-EDX, (c) STEM and (d) TEM-EDX images of
Textural and chemical characterization CTi40.

The actual TiO2 contents in the nal carbonized composites


were measured by TGA (Table 1). These weight percentages were
Fig. 2a and S1† show the N2 adsorption–desorption
slightly higher than the theoretical values because the weight
isotherms of the pure carbon and CTiX composites, which were
lost during carbonization was not exactly 50%; however, the
all of type I, typical of microporous solids,46 although there was
experimental values were similar to the expected values.
an increase in N2 uptake with a rise in relative pressure aer the
The morphology of the samples was studied by scanning
micropores were lled, which is indicative of the presence of
electron microscopy (Fig. 1). The CTiX samples (Fig. 1a) dis-
narrow mesopores with F z 4 nm. All isotherms also exhibited
played a typical structure of carbon xerogels and consisted of
a large increase in the amount adsorbed at a relative pressure of
primary nanospheres that were interconnected to form a three-
z0.95 owing to the condensation of N2 within void spaces. The
dimensional porous structure. A homogeneous structure was
BET surface area and porosity properties of composites are
obtained without differentiation between the TiO2 and carbon
important parameters that govern the performance of super-
phases. An HRTEM image (Fig. 1c) shows very high dispersion
capacitors. Details of the BET surface area and porosity prop-
of TiO2 nanoparticles in the carbon matrix, which was corrob-
erties are listed in Table 1.
orated by EDX analysis of these samples (Fig. 1b and d). Hence,
The results obtained from the adsorption of CO2 and N2
a homogeneous distribution of Ti in all the samples was ach-
show a gradual variation in the textural properties between the
ieved independently of the TiO2 loading present in each
composites as the percentage of TiO2 increased. On an increase
composite.
in the percentage of TiO2, the total porosity (V0.95) increased,
The hydrophobicity of the samples (Table 1) shows that the
which was favored by a growth in the volume of mesopores
TiO2–xerogel composites exhibited greater attraction to water
(Vmes) (Fig. 2b).
molecules, which greatly favoured the wettability of the carbon
The micropore distribution can be determined by comparing
phase and would increase the mobility of H+ and SO42 ions
the results of CO2 and N2 adsorption experiments. It is well
through the porous network. The hydrophobicity decreased as
known that the adsorption of CO2 provides information about
the TiO2 content increased owing to the presence of the inor-
narrow microporosity, i.e., ultramicropores, which correspond
ganic phase.

Table 1 Textural characteristics of the CTiX composites and pure phasesa

Sample TiO2% HF SBET m2 g1 W0(N2) cm3 g1 W0(CO2) cm3 g1 L0(N2) nm L0(CO2) nm Vmes cm3 g1 V0.95 cm3 g1

C100 0 0.21 585 0.228 0.275 0.58 0.56 0.088 0.316


CTi10 17 0.56 539 0.212 0.223 0.67 0.58 0.149 0.361
CTi20 22 0.75 479 0.189 0.205 0.75 0.56 0.187 0.376
CTi30 34 0.84 481 0.192 0.176 0.97 0.58 0.242 0.434
CTi40 45 0.96 423 0.157 0.174 1.03 0.59 0.323 0.480
TiO2 100 — 116 0.047 n.d. 1.89 n.d. 0.279 0.326
a
n.d. Not detected.

636 | J. Mater. Chem. A, 2018, 6, 633–644 This journal is © The Royal Society of Chemistry 2018
View Article Online

Paper Journal of Materials Chemistry A


Published on 27 November 2017. Downloaded by Ege Universitesi on 8/21/2020 11:33:47 AM.

Fig. 2 (a) N2 adsorption isotherms at 196  C on samples of C ( ), Fig. 3Evolution of pore volumes and mean pore widths with TiO2
CTi10 ( ), CTi20 ( ), CTi30 ( ) and CTi40 ( ); (b) variation of the values content: (a) micropore volume W0; (b) mean pore width L0. ( ) data
of Vmes ( ) and V0.95 ( ) with the TiO2 content. from N2 adsorption, ( ) data from CO2 adsorption.

to micropores with a diameter of less than 0.7 nm, whereas the


that the carbon matrix prevented the sintering of titania
total microporosity including supermicropores is determined
nanoparticles and stabilized the anatase form even at high
from the N2 isotherm only in the absence of diffusion restric-
temperatures.
tions. Fig. 3a shows the variations in the volumes of ultra- and
The surface chemistry of the titanium oxide composites was
supermicropores with the variations in the TiO2 content. The
analysed by XPS. The results are shown in Table 2 and Fig. S2.†
micropore volume (measured by N2 adsorption) decreased and
The deconvolution of the Ti 2p spectral region for TiO2 required
the pores became wider (L0(N2) increased) (Table 1) with an
only one peak at a binding energy (BE) of 459.2 eV, which cor-
increase in the titanium oxide content, whereas the ultra-
responded to Ti4+ and was in agreement with previously pub-
microporosity (W0(CO2)) progressively decreased and the mean
lished BE values,48 whereas two components were needed for
pore size was practically unaffected and only a slight increase
the CTiX samples at binding energies (BE) of 459 eV, which
was detected (Fig. 3b).
corresponded to Ti4+, and 458.2 eV, which was assigned to Ti3+
For the sample of C100 W0(N2) < W0(CO2), which indicated
species. Note that no partially reduced oxides were observed in
restrictions on the diffusion of N2 into micropores induced by
the pure TiO2 phase, whereas approximately 55% of the Ti
the narrow microporosity of the carbon phase. In contrast, for
atoms were reduced during the carbonization of the organic
the carbon–TiO2 composites W0(N2) z W0(CO2) as a conse-
gels.
quence of the abovementioned widening of the pores. In this
Two components at 530.8 and 532.2 eV, respectively, were
sense, the micropore width (L0) also increased progressively and
used to t the O 1s spectral region for TiO2. The major
consequently upon the widening of the micropores the above-
component of the O 1s spectral region corresponded to ‘bulk’
mentioned restriction on diffusion decreased. The mesopore
oxygen atoms in the stoichiometric form of TiO2, whereas the
volume increased more or less linearly with the TiO2 content
high-BE component could correspond to various oxygen-
(Fig. 2b), whereas the micropore volume (determined by N2 or
containing surface functional groups.49 Primarily, hydroxyl
CO2 adsorption) exhibited the reverse tendency (Fig. 3a).
(–OH) or bridging surface oxygen (Ti–O–Ti) groups were
Powder XRD measurements of CTiX samples were per-
described.50 However, four components were used to t the O 1s
formed to identify the crystalline domains of the embedded
spectral region for the samples of the composites, which was
titanium. As is well-known, TiO2 can exist as two crystalline
also inuenced by the content of oxygen bonded to the carbon
forms, namely, anatase and rutile. Anatase is a metastable
phase. The components at 530.2 eV and 531.3 eV corresponded
polymorphic form that can be transformed by heating to rutile.
For pure oxides, this transformation is fast at temperatures of
above 730  C, but the phase transition temperature has been
shown to depend on the impurity content, particle size, and
surface area.47 Fig. 4 shows the XRD patterns obtained for the
series of CTiX composites. At a TiO2 content of less than 20%,
the XRD patterns do not show any dened diffraction peak that
corresponds to the inorganic phase. At a TiO2 content of 30%
the observed peaks correspond to the anatase polymorphic
form. The above results indicate that in the carbonized
composite xerogels the metal oxide phase was well dispersed in
the carbon phase with a crystal size of less than 4 nm. Because
the carbonization temperature of the samples was 900  C, only
the rutile phase was expected in the nal carbon–titanium oxide
composites. Nevertheless, the anatase phase was observed in all Fig. 4 Powder XRD patterns of samples: CTi10 ( ), CTi20 ( ), CTi30
the carbon–titania composite materials. These results indicate ( ), CTi40 ( ) and TiO2 ( ).

This journal is © The Royal Society of Chemistry 2018 J. Mater. Chem. A, 2018, 6, 633–644 | 637
View Article Online

Journal of Materials Chemistry A Paper

Table 2 XPS results obtained from the deconvolution of high-resolution XPS spectra of C 1s, O 1s and Ti 2p

C 1s Ti 2p3/2 O 1s

Sample Assignment BE (eV) Peak (%) Assignment BE (eV) Peak (%) Assignment BE (eV) Peak (%) OXPS (wt%) TiXPS (wt%)

TiO2 — — — Ti4+ 459.2 100 Ti4+–O 530.8 92


–OH/Ti–O–Ti 532.2 8
C100 C]C 284.6 74 — — — C]O 531.6 18 2.65 —
C–O 285.6 13 C–O 533.2 82
C]O 286.6 5
COO– 287.5 3
Published on 27 November 2017. Downloaded by Ege Universitesi on 8/21/2020 11:33:47 AM.

CO2 289.3 3
p–p* 290.9 1
CTi10 C]C 284.6 64 Ti3+ 458.3 54 Ti4+–O 530.4 4 3.17 0.61
C–O 285.5 19 Ti4+ 459.2 46 Ti3+–O 531.3 2
C]O 286.6 7 C]O 532.5 53
COO– 287.5 3 C–O 533.9 41
CO2 289.2 5
p–p* 290.8 2
CTi20 C]C 284.6 73 Ti3+ 458.2 59 Ti4+–O 530.0 23 3.99 1.41
C–O 285.6 14 Ti4+ 458.7 41 Ti3+–O 531.3 12
C]O 286.6 5 C]O 532.6 36
COO– 287.6 2 C–O 533.9 29
CO2 289.1 3
p–p* 290.6 2
CTi30 C]C 284.6 71 Ti3+ 458.4 54 Ti4+–O 530.1 62 6.87 10.46
C–O 285.6 14 Ti4+ 459.0 46 Ti3+–O 531.3 14
C]O 286.7 6 C]O 532.7 16
COO– 287.8 3 C–O 533.9 7
CO2 289.5 4
p–p* 291.1 2
CTi40 C]C 284.6 66 Ti3+ 458.7 55 Ti4+–O 530.2 67 11.75 14.96
C–O 285.5 19 Ti4+ 459.2 45 Ti3+–O 531.1 22
C]O 287.0 7 C]O 532.2 6
COO– 288.6 3 C–O 533.0 5
CO2 289.9 2
p–p* 291.1 3

to ‘bulk’ oxygen atoms in the stoichiometric form of TiO2 and curves are observed for all samples except pure carbon, which
oxygen atoms bonded to Ti3+, respectively,49 whereas the peaks indicates pure capacitive behaviour.51 Data for CCV values ob-
at 532.6 and 533.9 eV were assigned to C]O and C–O groups in tained from these curves are presented in Fig. 5b. All samples
the carbon phase. The two latter peaks were predominant for displayed good capacitance retention with variations in the scan
the composites with low TiO2 contents, but with an increase in rate. Higher capacitance values were obtained using the
this parameter the main components were those corresponding composite materials in comparison with the pure carbon phase
to Ti4+ and Ti3+ species (see the marked change in the O although the surface area of the former was similar or even
distribution in Table 2). Note also that the BE values of the Ti lower. This behaviour indicates that the pore structure, as well
and O peaks were around 0.7 eV higher than those detected for as the chemical surface, played an important role. In that sense,
pure titania, which may be due to the ability of the carbon phase CTi10 and C100 had very similar microporous structures, in
to accommodate electrons from the inorganic phase, which led which the main difference was the higher mesopore volume in
to an electron-poor environment. CTi10, but these samples displayed very different capacitive
It is important to note the very low surface concentration of behaviour (CCV ¼ 42 vs. 97 F g1 for C100 and CTi10, respec-
Ti, in particular in the samples of CTi10 and CTi20. This fact is tively). This should be related to the presence of TiO2 on the
indicative of the high dispersion of TiO2 in the carbon matrix, surface. TiO2 could improve the capacitive behaviour by (i)
and TiO2 can also be embedded in the matrix and not detected improving the hydrophilicity of the surface, which signicantly
by XPS. increased the contact of the electrolyte with the surface and
favoured the diffusion of ions and (ii) the reduced species Ti3+
could act as an electron transfer centre and thus decrease the
Electrochemical characterization resistance of the electrode. Usually, an increase in the scan rate
Fig. 5a shows cyclic voltammograms (CV) recorded for pure induces a decrease in the capacitance, but in the case of the
carbon and TiO2–carbon composites with different ratios of xerogel–TiO2 samples this decrease was less than for the carbon
titanium oxide at a scan rate of 20 mV s1. Quasi-rectangular CV electrodes. In fact, the capacity retention between 0.5 and

638 | J. Mater. Chem. A, 2018, 6, 633–644 This journal is © The Royal Society of Chemistry 2018
View Article Online

Paper Journal of Materials Chemistry A

the composites in a three-electrode system using an Ag/AgCl


electrode as the reference electrode. The results (Fig. 5c) show
that the presence of TiO2 produced pseudofaradaic effects
owing to load storage, which were more signicant for the
composite with a TiO2 content of 20%.
In comparison with other TiO2–carbon nanomaterial (CNT
or graphene) composites (Table 3), the capacitance values in the
present work were much higher than the literature values (30–
300% higher), although a three-electrode system instead of
a symmetric supercapacitor was employed in the present work,
Published on 27 November 2017. Downloaded by Ege Universitesi on 8/21/2020 11:33:47 AM.

where the pseudocapacitive contribution was not taken into


account. Despite this effect, the capacitance values were of the
same order as or even greater than those of asymmetric
supercapacitors.
One limiting factor on the use of carbon materials as elec-
trodes for supercapacitors is their poor capacitance retention at
high current densities. As can be seen in Fig. 6b, in the TiO2–
carbon xerogel composites this reduction in capacitance was
lower than in C100. This fact must be related to the restrictions
on ion diffusion in the pore network. These restrictions are
Fig. 5 (a) Cyclic voltammograms recorded in 1 M H2SO4 at a scan rate controlled by two factors: the rst is the pore size in the
of 20 mV s1. (b) Variation of the gravimetric capacitance CCV with the micropore network and the second is the wettability of the
scan rate. (c) Cyclic voltammograms recorded at 5 mV s1 for the carbon surface. Thus, an adequate combination of both factors
composites in a three-electrode system. Samples: C100 ( ), CTi10 ( ), is necessary in order to obtain high capacitance and high
CTi20 ( ), CTi30 ( ), and CTi40 ( ).
capacitance retention. Fig. 7a shows the linear variation in
capacitance retention with hydrophobicity for all samples that
were studied, which indicates the great inuence of this
30 mV s1 was 74.2%, 78.4%, 73.5% and 75.6% for CTi10, CTi20, parameter on the efficiency of the electrodes. Fig. 7b shows the
CTi30 and CTi40, respectively. Moreover, literature results (Table variation in capacitance retention with the micropore width
3) show that the incorporation of Ti3+ sites by self-doping determined from the nitrogen adsorption isotherm. In this
processes (plasma treatment or hydrogenation reactions) will case, for a micropore size of greater than 0.75 nm the retention
generate abundant disordered states and lead to a negative effect was practically constant.
on cycling stability. Lu et al.17 showed a decrease in capacity The electrical conductivities of the samples were determined
retention when the hydrogenation temperature was increased, by electrochemical impedance spectroscopy, and the Nyquist
which could be ascribed to an increase in the number of Ti3+ plots are shown in Fig. 8a. The equivalent series resistance
sites. In our case, although a large proportion of Ti3+ sites was (ESR) is obtained from the intersection of the impedance with
detected by XPS (around 55%), high capacity retention was the real axis at high frequency.52 The diameter of the semicircle
observed, which was even higher than that of other carbon–TiO2 corresponds to the so-called charge transfer resistance RCT53 at
composites (Table 3) used in the literature. the electrode–electrolyte interface, which is oen correlated
Galvanostatic charge–discharge curves of the samples recor- with the electronic conductivity of the electrode. The results
ded at 2 A g1 are shown in Fig. 6a. The shapes of these curves for that were obtained are listed in Table 4.
all samples are almost ideally triangular. As determined by CV, In the high-frequency region the ESR values are obtained.
the capacitance of the sample of CTi20 was higher than those of The ESR represents the sum of the intrinsic resistance of the
the other samples owing to a combination of two factors, namely, active material, the resistance of the electrolyte, and the contact
its optimal textural characteristics and low hydrophobicity (Table resistance between the carbon material electrode and the
1). The low hydrophobicity of the TiO2–carbon xerogels improved current collector. The ESR values shown in Table 4 approxi-
the wettability of the carbon surface and could facilitate the mately reect the comparative conductive properties of the
formation of an EDL, which favored the mobility of ions during activated carbon electrodes, because in all cases the same
the electrochemical measurements. This should be the reason electrolyte, collectors, and type of cell were used to assemble the
why high capacitance values were observed at high current rates, supercapacitor.52 The ESR values were approximately identical
ranging from 54 to 81 F g1 at 20 A g1. Fig. 6b shows the vari- for all the carbon xerogel–TiO2 composites and were much
ation of the capacitance with the current rate. It is clear that all lower than in the case of pure carbon.
samples exhibited a decrease in capacitance when the current The samples of CTi20 and CTi30 exhibited the lowest RCT
rate increased from 62 mA g1 to 20 A g1, which was due to values of 0.70 and 0.76 U, respectively. The samples of CTi10
restrictions on the formation of an EDL at high current rates. and CTi40 displayed higher charge transfer resistances (1.25
To determine possible pseudofaradaic contributions of the and 1.86 U, respectively), and the sample of C100 exhibited the
TiO2 phase, cyclic voltammetry was performed at 5 mV s1 on highest value of all the samples. These results clearly show that

This journal is © The Royal Society of Chemistry 2018 J. Mater. Chem. A, 2018, 6, 633–644 | 639
Published on 27 November 2017. Downloaded by Ege Universitesi on 8/21/2020 11:33:47 AM.

Table 3 Electrochemical performance of carbon–TiO2 composites from the literature

C retention (%)

TiO2 loading At 30 Cmax Cmax Emax Pmax


Sample (wt%) 5000 cycles mV s1 (F g1) (mF cm2) (W h kg1) (W kg1) Capacitance measurement Electrolyte Cell type Ref.

TiO2–CNT 0 — — 7 (—) — — — 10 mV s1 (0.05 mA cm2) 1 M H2SO4 Three-electrode 24


6.3 — — 14 (90) — — —
17.4 90 (200 cycl.) — 24 (110) — 55 580
33.0 — — 43 (45) — — —
Journal of Materials Chemistry A

17.4 — — 42 — — — 2.5 mA cm2


TiO2– 0 — — 25 (15) — — — 10 mV s1 (150 mV s1) 1 M KOH Three-electrode 23

640 | J. Mater. Chem. A, 2018, 6, 633–644


graphene 57.4 — 58 75 (36.7)
73.6 88 (1000 cycl.) 57 84 (43)
TiO2/C 74.5 97.2 18.4 — 1.60 69 500 mA g1 1 M KOH Three-electrode 22
85.7 97.4 13.0 1.10 69
92.0 98.1 10.5 0.70 69
0 90 16.0 1.05 69
TiO2 — — — 4.1 — — — 20 mV s1 1 M H2SO4 Two-electrode 34
TiO2–CNT — — — 36.8 — — —
C/TiO2/rGO — 97 (500 cycl.) — — 23.6 — — 50 mV s1 0.5 M Na2SO4 Two-electrode 35
C–TiO2 0 100 23.5 23.8 (42.5) — — — 10 mV s1 (0.5 mV s1) 1 M H2SO4 Two-electrode our
10 100 74.2 77.0 (97) — — — work
20 100 78.4 95.0 (109) — 20.7 20.0 69 600
30 100 73.5 76.0 (93) — — —
40 100 75.6 77.0 (89) — — —
TiO2 — — 40a 32 (63) — — — 10 mV s1 (2 mV s1) 1 M LiPF6 Two-electrode 28
(commercial)
TiO2 — 64a 52 (70) — — —
(synthesized)
TiO2/rGO — 61a 63 (90) — 42 800
C — — — 162.2 (75) — — — 10 mV s1 (400 mV s1) 5 M LiCl Three-electrode 26
TiO2@C — — — 197.1 (100) — — —
H–TiO2@C — — — 253.4 (175) — — —
MnO2 — — — 325.8 (100) — — — 10 mV s1 (200 mV s1)
TiO2@MnO2 — — — 359.7 (150) — — —
H– — — — 449.6 (250) — — —
TiO2@MnO2
H– — 82 70 141.8 — 59 45 kW kg1 10 mV s1 Two-electrode ASC
TiO2@MnO2//
H–TiO2@C
SiO2@C/TiO2 — — — 1018 — — — 1 A g1 1 M KOH Three-electrode 29
hollow — 74.4 — 73.4 29 375 Two-electrode ASC
spheres
Mn-doped — 88.1 57.3 — 169.5 (71.5) — — 10 mV s1 (100 mV s1) 0.5 M Na2SO4 Three-electrode 19
TiO2 porous
lm
View Article Online

This journal is © The Royal Society of Chemistry 2018


Paper
View Article Online

Paper Journal of Materials Chemistry A

Ref.

17
Three-electrode
Cell type

0.5 M Na2SO4
Published on 27 November 2017. Downloaded by Ege Universitesi on 8/21/2020 11:33:47 AM.

Electrolyte

Fig. 6 (a) Chronoamperometric profiles recorded in 1 M H2SO4 at


a current rate of 2 A g1. (b) Variation of the gravimetric capacitance
CGD with the current rate. Samples: C100 ( ), CTi10 ( ), CTi20 ( ),
CTi30 ( ), and CTi40 ( ).
Capacitance measurement

the presence of TiO2 in the carbon material induced a decrease


in the resistance of the electrode and improved its performance
for energy storage. These trends should be correlated with the
100 mV s1

pore structure and conductivity of the sample. Fig. 8b shows the


correlation between the ESR and RCT resistances and the
micropore width; the minimum value of the charge transfer
resistance corresponds to a width of 0.75 nm, which indicates
that this is the most suitable pore diameter for energy storage in
(W kg1)

this kind of material. In fact, the sample of CTi20 displayed


Pmax

a higher capacitance and lower RCT resistance with a pore







diameter of 0.75 nm as determined by N2 adsorption.


(W h kg1)

The optimal micropore diameter determined in this work is


slightly larger than that calculated by other authors52,54 and is
Emax

larger than the sizes of hydronium ions (0.36–0.42 nm)54 and







hydrated bisulfate ions (0.53 nm),32 but in this case the elec-
trodes contained a large amount of inorganic material, which
(mF cm2)

0.33 (0.21)

increased the interaction between ions and electrodes.


z0.08

0.026
z3.5
z6.2

Fig. 9 shows the variations in the imaginary part of the


Cmax

3.24

0.08

capacitance (C00 ) with the frequency and indicates the transi-


tional frequency between pure capacitive and pure resistive
behavior. The relaxation time constant (s) is obtained from the
(F g1)

frequency f0 at the maximum of the curve by the expression s ¼


Cmax

1/(2p  f0)55 and is a quantitative measure of the speed of







discharge of the device.56 The results in Table 4 show that the s


mV s1
At 30

75.0
63.3
43.3



C retention (%)

5000 cycles

96.9

55.7
65.2



TiO2 loading

Retention at 20 mV s1.
(wt%)







(Contd. )

H–TiO2 300
400
500
600

Fig. 7 Variation of retention of capacitance CGD (100  7 A g1/


Mn-doped

nanotube

Air–TiO2

0.125 A g1) for carbon and carbon xerogels with (a) hydrophobicity
Sample
Table 3

arrays
TiO2

TiO2

and (b) mean micropore size determined by N2 adsorption. Samples:


C100 ( ), CTi10 ( ), CTi20 ( ), CTi30 ( ), and CTi40 ( ).
a

This journal is © The Royal Society of Chemistry 2018 J. Mater. Chem. A, 2018, 6, 633–644 | 641
View Article Online

Journal of Materials Chemistry A Paper


Published on 27 November 2017. Downloaded by Ege Universitesi on 8/21/2020 11:33:47 AM.

Fig. 8 (a) Nyquist plots and (b) relationship of ESR (closed symbols) Fig. 10 Ragone plots for samples: C100 ( ), CTi10 ( ), CTi20 ( ),
and RCT (open symbols) with mean micropore width. Samples: C100 CTi30 ( ), and CTi40 ( ).
( ), CTi10 ( ), CTi20 ( ), CTi30 ( ), and CTi40 ( ).

Table 5 Maximum (at 69 W kg1) and minimum (at 4500 W kg1)


Table 4 CEIS at 1 mHz, equivalent series resistance (ESR), charge energy densities (W h kg1) from Ragone plots
transfer resistance (RCT) and relaxation time constant (s) obtained from
EIS Sample Emax W h kg1 Emin W h kg1 ERet.%

Sample CEIS (F g1) ESR (U) RCT (U) s (s) C100 4.3 0.5 11.6
CTi10 16.9 7.9 46.7
C100 59 0.60 3.46 11.44 CTi20 20.7 13.5 65.2
CTi10 102 0.46 1.25 0.26 CTi30 17.7 8.3 46.9
CTi20 135 0.48 0.70 0.14 CTi40 15.0 6.8 45.3
CTi30 101 0.48 0.76 0.19
CTi40 94 0.43 1.86 0.26

a supercapacitor based on CTi20 was greater than that of other


materials used for commercially available supercapacitors
values for the carbon composite materials were lower than that
(usually less than 10 W h kg1).1 The energy that was released
for pure carbon and the lowest value corresponded to CTi20,
decreased at higher power densities (Table 5). Thus, the energy
which showed the excellent performance of this material as an
density retention reached a maximum for CTi20, for which the
electrode for energy storage and release.
energy released at 4.5 kW kg1 was 65.2% of that released at
Ragone plots for the carbon xerogel–TiO2 composites are
69 W kg1.
presented in Fig. 10 and show the relationship between the
The stability of electrodes during charge–discharge cycles is
energy density and power density. Table 5 shows the results for
a key factor for their practical application. In this work the
all the samples that were studied for the maximum energy
stability was evaluated by galvanostatic charge–discharge
density and energy density at a power density of 4500 W kg1,
cycling. Fig. 11 shows the results of cycling stability tests for the
which was the maximum value common to all samples. The
samples studied in this work.
maximum energy density drastically increased upon doping
The capacitance of the samples remained practically
with TiO2 owing to the better wettability of the sample and the
unchanged aer 13 000 charge–discharge cycles at 1 A g1. The
widening of the micropores. The maximum value reached was
resistance was practically constant throughout the cycles except
20.7 W h kg1 for the sample of CTi20. The energy density of
for the samples of C100 and CTi40, which exhibited a slight

Fig. 11 Galvanostatic charge–discharge cycles at a current rate of


Fig. 9 Evolution of imaginary part of capacitance vs. frequency. 1 A g1: C100, ( ); CTi10, ( ); CTi20, ( ); CTi30, ( ); CTi40, ( ).
Samples: C100 ( ), CTi10 ( ), CTi20 ( ), CTi30 ( ), and CTi40 ( ). Capacitance (open symbols), ohmic resistance (closed symbols).

642 | J. Mater. Chem. A, 2018, 6, 633–644 This journal is © The Royal Society of Chemistry 2018
View Article Online

Paper Journal of Materials Chemistry A

increase in resistance aer 6500 cycles, which indicated a slight Andalucı́a). A. Elmouwahidi acknowledges a predoctoral
degradation of the electrodes. The results for these carbon fellowship from the Erasmus Mundus Al-Idrisi, programme.
composites indicated high stability in potential uses as elec-
trodes for supercapacitors, in particular in the case of CTi20. References
Conclusions 1 G. Wang, L. Zhang and J. Zhang, Chem. Soc. Rev., 2012, 41,
797–828.
Carbon xerogel–TiO2 composites (CTiX) were successfully 2 P. Simon and Y. Gogotsi, Nat. Mater., 2008, 7, 845–854.
synthesized by sol–gel techniques. The composites had 3 T. Chen and L. Dai, Mater. Today, 2013, 16, 272–280.
a homogeneous three-dimensional mesoporous structure, in 4 B. E. Conway, Similarities and Differences between
Published on 27 November 2017. Downloaded by Ege Universitesi on 8/21/2020 11:33:47 AM.

which both phases were also homogeneously and intimately Supercapacitors and Batteries for Storing Electrical Energy,
distributed. All the textural, chemical, crystallographic and Springer US, Boston, MA, 1999.
catalytic properties were determined by the TiO2 content in the 5 M. Zhi, C. Xiang, J. Li, M. Li and N. Wu, Nanoscale, 2013, 5,
composite and by the synergetic effect between both phases. 72–88.
Carbon xerogels are predominantly microporous supports, but 6 T. Lu, Y. Zhang, H. Li, L. Pan, Y. Li and Z. Sun, Electrochim.
the composites were micro/mesoporous materials with an Acta, 2010, 55, 4170–4173.
increase in mesoporosity at the expense of microporosity. The 7 X. Sun, M. Xie, G. Wang, H. Sun, A. S. Cavanagh, J. J. Travis,
presence of carbon in the composite prevented the crystal S. M. George and J. Lian, J. Electrochem. Soc., 2012, 159,
growth of TiO2 and hindered the transition from anatase to A364–A369.
rutile. Owing to the interactions between both phases all the 8 R. Górniak and S. Lamperski, J. Phys. Chem. C, 2014, 118,
composites exhibited a high dispersion of anatase TiO2 nano- 3156–3161.
particles on the carbon support even though the samples were 9 V. D. Patake, C. D. Lokhande and O. S. Joo, Appl. Surf. Sci.,
carbonized at 900  C. These nanoparticles were partially 2009, 255, 4192–4196.
reduced with a surface distribution of oxygen that was highly 10 D. Yan, Z. Guo, G. Zhu, Z. Yu, H. Xu and A. Yu, J. Power
dependent on the TiO2 content. Sources, 2012, 199, 409–412.
The supercapacitor performance of the samples was analysed 11 U. M. Patil, R. R. Salunkhe, K. V. Gurav and C. D. Lokhande,
in a two-electrode system. The composites exhibited improved Appl. Surf. Sci., 2008, 255, 2603–2607.
electrochemical performance with high capacitance (up to 12 M. S. Kim, T. W. Lee and J. H. Park, J. Electrochem. Soc., 2009,
137 F g1 at 0.250 A g1 for the composite containing 20% TiO2), 156, A584–A588.
high capacitance retention (66–80%) at 20 A g1 and a high 13 M. Zhou, A. M. Glushenkov, O. Kartachova, Y. Li and
energy density of 20.7 W h kg1 at a power density of 69 W kg1 in Y. Chen, J. Electrochem. Soc., 2015, 162, A5065–A5069.
the voltage range of 0 V to 1.1 V, which was higher than those of 14 M. Salari, S. H. Aboutalebi, A. T. Chidembo, I. P. Nevirkovets,
most reported TiO2–C composites. This improved behaviour K. Konstantinov and H. K. Liu, Phys. Chem. Chem. Phys.,
could be related to different properties of the composites: (i) the 2012, 14, 4770–4779.
presence of TiO2 improved the hydrophilicity of the surface, 15 H. Wu, C. Xu, J. Xu, L. Lu, Z. Fan, X. Chen, Y. Song and D. Li,
which signicantly increased the contact of the electrolyte with Nanotechnol., 2013, 24, 455401.
the surface and favoured the diffusion of ions; (ii) the carbon 16 B. Chen, J. Hou and K. Lu, Langmuir, 2013, 29, 5911–5919.
matrix favoured the reduction of TiO2, and the reduced species 17 X. Lu, G. Wang, T. Zhai, M. Yu, J. Gan, Y. Tong and Y. Li,
Ti3+ could act as an electron transfer centre and thus decrease the Nano Lett., 2012, 12, 1690–1696.
resistance of the electrode; (iii) there were pseudofaradaic 18 H. Wang, Z. Tang, L. Sun, Y. He, Y. Wu and Z. Li, Rare Met.,
contributions from the TiO2 phase; and (iv) the presence of TiO2 2009, 28, 231–236.
in the carbon material induced a decrease in the resistance of the 19 X. Ning, X. Wang, X. Yu, J. Zhao, M. Wang, H. Li and Y. Yang,
electrodes, which improved their performance for energy storage Sci. Rep., 2016, 6, 22634.
owing to optimization of the pore structure. A sample with 20 H. Li, Z. Chen, C. K. Tsang, Z. Li, X. Ran, C. Lee, B. Nie,
a combination of low hydrophobicity and an adequate micro/ L. Zheng, T. Hung, J. Lu, B. Pan and Y. Y. Li, J. Mater.
mesopore network with an intermediate content of TiO2 dis- Chem. A, 2014, 2, 229–236.
played the best performance for energy storage. A oating test 21 G. D. Moon, J. B. Joo, M. Dahl, H. Jung and Y. Yin, Adv. Funct.
showed the very good cyclability of the synthesized materials. Mater., 2014, 24, 848–856.
22 L. Lu, Y. Zhu, F. Li, W. Zhuang, K. Y. Chan and X. Lu, J.
Conflicts of interest Mater. Chem., 2010, 20, 7645–7651.
23 X. Sun, M. Xie, G. Wang, H. Sun, A. S. Cavanagh, J. J. Travis,
There are no conicts to declare. S. M. George and J. Lian, J. Electrochem. Soc., 2012, 159,
A364–A369.
Acknowledgements 24 C. T. Hsieh, C. C. Chang, W. Y. Chen and W. M. Hung, J.
Phys. Chem. Solids, 2009, 70, 916–921.
This research is supported by the FEDER and Spanish projects 25 S. Yang, Y. Lin, X. Song, P. Zhang and L. Gao, ACS Appl.
CTQ2013-44789-R (MINECO) and P12-RNM-2892 (Junta de Mater. Interfaces, 2015, 7, 17884–17892.

This journal is © The Royal Society of Chemistry 2018 J. Mater. Chem. A, 2018, 6, 633–644 | 643
View Article Online

Journal of Materials Chemistry A Paper

26 X. Lu, M. Yu, G. Wang, T. Zhai, S. Xie, Y. Ling, Y. Tong and 40 L. Bonnefoi, P. Simon, J. F. Fauvarque, C. Sarrazin,
Y. Li, Adv. Mater., 2013, 25, 267–272. J. F. Sarrau and A. Dugast, J. Power Sources, 1999, 80, 149–
27 A. Ramadoss and S. J. Kim, Electrochim. Acta, 2014, 136, 105– 155.
111. 41 L. L. Zhang and X. S. Zhao, Chem. Soc. Rev., 2009, 38, 2520–
28 H. Kim, M. Y. Cho, M. H. Kim, K. Y. Park, H. Gwon, Y. Lee, 2531.
K. C. Roh and K. Kang, Adv. Energy Mater., 2013, 3, 1500– 42 D. Hulicova-Jurcakova, M. Seredych, G. Q. Lu and
1506. T. J. Bandosz, Adv. Funct. Mater., 2009, 19, 438–447.
29 Y. Zhang, Y. Zhao, S. Cao, Z. Yin, L. Cheng and L. Wu, ACS 43 T. E. Rufford, D. Hulicova-Jurcakova, E. Fiset, Z. Zhu and
Appl. Mater. Interfaces, 2017, 9, 29982–29991. G. Q. Lu, Electrochem. Commun., 2009, 11, 974–977.
30 R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki and Y. Taga, 44 G. A. Ferrero, A. B. Fuertes and M. Sevilla, Electrochim. Acta,
Published on 27 November 2017. Downloaded by Ege Universitesi on 8/21/2020 11:33:47 AM.

Science, 2001, 293, 269. 2015, 168, 320–329.


31 M. Salari, S. H. Aboutalebi, A. T. Chidembo, I. P. Nevirkovets, 45 T. Liang, C. Chen, X. Li and J. Zhang, Langmuir, 2016, 32,
K. Konstantinov and H. K. Liu, Phys. Chem. Chem. Phys., 8042–8049.
2012, 14, 4770–4779. 46 M. Thommes, K. Kaneko, A. V. Neimark, J. P. Olivier,
32 E. Bailón-Garcı́a, F. J. Maldonado-Hódar, F. Carrasco-Marı́n F. Rodriguez-Reinoso, J. Rouquerol and S. W. Sing
and A. F. Pérez-Cadenas, Carbon-gel based catalysts, Kenneth, Pure Appl. Chem., 2015, 87, 1051–1069.
WO2016/174295 A2, 2016. 47 D. A. H. Hanaor and C. C. Sorrell, J. Mater. Sci., 2011, 46, 855–
33 E. Bailón-Garcı́a, F. J. Maldonado-Hódar, F. Carrasco-Marı́n 874.
and A. F. Pérez-Cadenas, Procedimiento de preparación de 48 A. R. Gonzalez-Elipe, P. Malet, J. P. Espinos, A. Caballero and
foto-catalizadores, foto-catalizadores obtenibles por el G. Munuera, Stud. Surf. Sci. Catal., 1989, 48, 427–436.
mismo y procedimiento de fotodegradación que los usa, 49 M. E. Nagassa, A. E. Daw, W. G. Rowe, A. Carley,
ES 2 538 627 B1, 2016. D. W. Thomas and R. Moseley, Clinical oral implants
34 C. J. Chien, S. S. Deora, P. Chang, D. Li and J. G. Lu, IEEE research, 2008, 19, 1317–1326.
Trans. Nanotechnol., 2011, 10, 706–709. 50 A. Samokhvalov, E. C. Duin, S. Nair and B. J. Tatarchuk, Surf.
35 Q. Ke, Y. Liao, S. Yao, L. Song and X. A. Xiong, J. Mater. Sci., Interface Anal., 2010, 42, 1476–1482.
2016, 51, 2008–2016. 51 M. Minakshi Sundaram, A. Biswal, D. Mitchell, R. Jones and
36 Q. Wang, Z. Wen and J. Li, J. Nanosci. Nanotechnol., 2007, 7, C. Fernandez, Phys. Chem. Chem. Phys., 2016, 18, 4711–4720.
3328–3331. 52 D. Hulicova, M. Kodama and H. Hatori, Chem. Mater., 2006,
37 R. P. Bansal, J. P. Donnet and F. Stoeckli, Active carbon, 18, 2318–2326.
Marcel Dekker, New York, 1988. 53 C. Delacourt, P. Ridgway, V. Srinivasan and V. Battaglia, J.
38 F. Carrasco-Marin, A. Mueden, A. Centeno, F. Stoeckli and Electrochem. Soc., 2014, 161, A1253–A1260.
C. Moreno-Castilla, J. Chem. Soc., Faraday Trans., 1997, 93, 54 C. Moreno-Castilla, M. B. Dawidziuk, F. Carrasco-Marı́n and
2211–2215. E. Morallón, Carbon, 2012, 50, 3324–3332.
39 J. Gamby, P. L. Taberna, P. Simon, J. F. Fauvarque and 55 K. Wang, Y. Song, R. Yan, N. Zhao, X. Tian, X. Li, Q. Guo and
M. Chesneau, J. Power Sources, 2001, 101, 109–116. Z. Liu, Appl. Surf. Sci., 2017, 394, 569–577.
56 E. Raymundo-Pinero and F. Béguin, Interface Sci. Technol.,
2006, 7, 293–343.

644 | J. Mater. Chem. A, 2018, 6, 633–644 This journal is © The Royal Society of Chemistry 2018

You might also like