Preparation of mesohollow and microporous carbon nanofiber and its application in cathode material for lithium-sulfur batteries

Article abstract of DOI:10.1016/j.jallcom.2014.04.073

Mesohollow and microporous carbon nanofibers (MhMpCFs) were prepared by a coaxial electrospinning with polyacrylonitrile (PAN) and polymethylmethacrylate (PMMA) as outer and inner spinning solutions followed by a carbonization. The carbon fibers were thermal treated with sublimed sulfur to form S/MhMpCFs composite, which was used as cathode material for lithium-sulfur batteries. Electrochemical study shows that the S/MhMpCFs cathode material provides a maximum capacity of 815 mA h/g after several cycles of activation, and the capacity retains 715 mA h/g after 70 cycles, corresponding to a retention of 88%. The electrochemical property of the S/MhMpCFs composite is much superior than the S-incorporated solid carbon fibers prepared from electrospinning of single PAN. The mechanism of the enhanced electrochemical property of the S/MhMpCFs composite is discussed.

Full text of DOI:10.1016/j.jallcom.2014.04.073

Journal of Alloys and Compounds 608 (2014) 220–228
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jalcom
Preparation of mesohollow and microporous carbon nanofiber
and its application in cathode material for lithium–sulfur batteries
⇑
⇑
Yuanhe Wu, Mingxia Gao , Xiang Li, Yongfeng Liu, Hongge Pan
State Key Laboratory of Silicon Materials, Key Laboratory of Advanced Materials and Applications for Batteries of Zhejiang Province & Department of Materials Science
and Engineering, Zhejiang University, Hangzhou 310027, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Mesohollow and microporous carbon nanofibers (MhMpCFs) were prepared by a coaxial electrospinning
with polyacrylonitrile (PAN) and polymethylmethacrylate (PMMA) as outer and inner spinning solutions
followed by a carbonization. The carbon fibers were thermal treated with sublimed sulfur to form
S/MhMpCFs composite, which was used as cathode material for lithium–sulfur batteries. Electrochemical
study shows that the S/MhMpCFs cathode material provides a maximum capacity of 815 mA h/g after
several cycles of activation, and the capacity retains 715 mA h/g after 70 cycles, corresponding to a reten-
tion of 88%. The electrochemical property of the S/MhMpCFs composite is much superior than the
S-incorporated solid carbon fibers prepared from electrospinning of single PAN. The mechanism of the
enhanced electrochemical property of the S/MhMpCFs composite is discussed.
Received 21 February 2014
Received in revised form 7 April 2014
Accepted 10 April 2014
Available online 21 April 2014
Keywords:
Lithium–sulfur batteries
Sulfur cathode
Electrospinning
Porous carbon fibers
Electrochemical performance
Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction
the intermediate products of lithium polysulfides (Li2Sn,
4 6 n < 8) in the organic electrolyte lowers the utilization of sulfur
With the development of high-performance portable electric
devices, electric vehicles and battery energy-storage systems, sec-
ondary batteries with high-energy density and high-power density
are considerably important. In terms of specific energy, lithium–
sulfur (Li–S) batteries are promising. Li–S batteries provide a theo-
retical specific energy of 2567 W h/kg, which is much higher than
387 W h/kg of today’s commercial LiCoO2/C lithium-ion batteries
(LIBs) [1]. As the cathode material of Li–S batteries, sulfur has a
theoretical specific capacity of 1675 mA h/g and an energy density
of 2600 W h/kg, assuming that sulfur is fully reduced to Li2S [2,3].
Furthermore, sulfur has other advantages, such as low cost, natural
abundance, and environmental friendliness [4]. Therefore, Li–S
batteries show great potential for the next generation secondary
lithium batteries and have been drawn much attention in recent
years. However, Li–S batteries also suffer from several problems
[5–8]. First, sulfur is electrically and ionically insulating, leading
to poor electrochemical accessibility and low utilization of sulfur
in the electrode, so it cannot serve as cathode solely. Second, the
reduction products of Li2S/Li2S2 are insulating. They precipitate
on the electrode surface and thus increase consequentially the
internal resistance of the battery. Third, the high solubility of
and hence results in poor cycle performance of the batteries.
The main approaches to overcome the insulation of S focus on
introducing electrical conductive polymers, such as polyaniline
[9,10], polypyrrole [11,12] and polythiophene [13], carbonaceous
materials, such as carbon nanotubes [14–16], porous carbon
[17–19], carbon spheres [20] and grapheme [21], and the combina-
tion of electrical conductive polymers and carbonaceous materials
[22–24]. In addition to promote the electrical conductivity of the S
cathode, many of the above approaches also provide simulta-
neously a secondary action in improving the adsorption or store
of sulfur by using porous electrical conductive matrices [17–19]
to alleviate the dissolution of the intermediate polysulfides during
cycling. Pore-structured electrical conductive matrices are appro-
priate in improving both the rate and capacity retention perfor-
mances of the sulfur electrodes [25–27]. Previous investigations
suggested that the ideal porous structure of a matrix for sulfur
loading should be a hierarchical one consisting of properly distrib-
uted meso- and micro-pores [27–29]. Micropores act as the host of
the impregnated sulfur and prevent the dissolution of polysulfides
while the mesopores construct paths for electrolyte and make it
easy for the sulfur to enter the deep pores. Furthermore, high
specific surface area of the electrical conductive matrices is
also important in increasing the utilization of S cathode, as it offers
more sites for the deposition of the insulating lithiation products of
Li2S/Li2S2 [30]. In addition, mesoporous metal–organic frameworks
⇑
Corresponding authors. Tel./fax: +86 571 87952615.
E-mail addresses: gaomx@zju.edu.cn (M. Gao), hgpan@zju.edu.cn (H. Pan).
http://dx.doi.org/10.1016/j.jallcom.2014.04.073
0925-8388/Ó 2014 Elsevier B.V. All rights reserved.

Y. Wu et al. / Journal of Alloys and Compounds 608 (2014) 220–228
221
in the glove box (Labstar) with a porous polyethylene sheet (Celgard 2400, Celgard)
as separator, and a mixture of 1,3-dioxolane (DOL) and dimethoxymethane (DME)
(1:1 by volume) with 2 wt.% LiNO3 addition as electrolyte. The electrochemical
properties of the S/MhMpCFs composite cathode was tested by galvanostatic dis-
charge (delithiation) and charge (lithiation) in a potential range of 1.5–3.0 V (vs.
Li+/Li) at a current density of 0.1 C (1 C = 1675 mA/g). Cyclic voltammetry (CV) mea-
surement was conducted in a potential range of 1.5–3.0 V (vs. Li+/Li) at a scanning
rate of 0.1 mV/s for first 3 cycles by using an electrochemical workstation (MATAT,
Arbin Instrument, USA). Electrochemical impedance spectrum (EIS) measurements
were performed using a frequency response analyzer (Solartron 1255B, Solartron)
equipped with an electrochemical interface (1287, Solartron) in a frequency range
of 100 kHz to 0.01 Hz and a potentiostatic signal amplitude of 5 mV. Prior to the EIS
measurement, the cell was activated by three cycles at 0.1 C between 1.5–3.0 V vs.
Li+/Li. All of the electrochemical tests were performed at 25 1 °C.
have also been investigated as host materials for S cathodes and
showed favorable electrochemical properties [31,32].
Electrospinning is an effective method to prepare conductive
carbon nanofibers, which have also been investigated in anode
materials [33–38] and cathode materials [39] for LIBs, Li–S batter-
ies [40]. Polyacrylonitrile (PAN) electrospinning fibers can be read-
ily carbonized to form carbon nanofibers [37,41]. Compared with
particle materials, one-dimensional carbon nanofiber material pos-
sesses commonly higher electrical conductivity. In addition, carbon
nanofibers have fine mechanical strength, which is also supposed
to have good resistance for the volume change of the combined
active materials during lithium insertion and extraction. Polymeth-
ylmethacrylate (PMMA) can produce volatile degradation products
and creating micropores in the carbonization process of PAN while
it is incorporated with PAN [40].
For comparison, uniaxial electrospinning by using only the PAN solution was
used to prepare solid carbon nanofibers, which were further incorporated with S
in a same weight ratio as the porous ones as cathode material for Li–S batteries. Par-
allel analyses on the structure and electrochemical properties of the S/solid carbon
nanofibers (S/SCFs) were also performed.
In the present study, PAN and PMMA were used as outer and
inner spinning solutions to prepare mesohollow and microporous
carbon nanofibers via a coaxial electrospinning method followed
by a carbonization. It was found that PMMA was eliminated during
the carbonization process of PAN and created mesohollows and
micropores in the carbon fibers. With further introducing S by
thermal treatment, the composite of S/mesohollow and micropo-
rous carbon nanofibers (S/MhMpCFs) shows superior electrochem-
ical property as cathode material for Li–S batteries, which is much
better than that of the composite of S-incorporated solid carbon
nanofibers prepared from electrospinning of single PAN. The syn-
thesis method of the MhMpCFs is facile, which is considered help-
ful in providing new hierarchically porous carbon matrix for high-
performance S cathode for Li–S batteries.
3. Results and discussion
3.1. Structural characterization
Fig. 1 shows the SEM micrographs of the as-spun webs of PAN/
PMMA. It is seen that continuous fibrous webs with smooth surface
and almost even diameters of 0.8–1 lm form by electrospinning.
The morphology of the carbonized webs (at 900 °C for 1 h) after
grinding is shown in Fig. 2a. It shows that hollow fibers with
outer-diameter in the range of 200–300 nm form. The length of
the fibers is in an order of several micrometers, which is not even,
due to the uncontrollable grinding. Comparison of Figs. 1 and 2a
show that the diameters of the fibers are much smaller that of
the organic webs, indicating that the fibers were condensed after
carbonization. The PMMA in the core of the as-spun webs was
decomposed to gaseous products and eliminated during the car-
bonization process, leaving the hollow cores in the fibers. The
diameter of the hollows is not even. It is in the range of 30–
60 nm as seen from Fig. 2a. Fig. 2b shows the morphology of the
as-ground fibers prepared from carbonizing single PAN spinning.
The fibers show solid structure with no hollow observed as
expected. The diameter of the fibers is in the range of 300–
400 nm, slightly larger than that of the hollow ones, which is sup-
posed due to the less shrinkage during carbonization without the
addition of PMMA.
Micropores cannot be observed under SEM for either type of the
carbon fibers. However, N2 adsorption isotherm measurement and
the obtained pore size distribution of the two types carbonized
fibers, which are shown in Fig. 3a and b, respectively, demonstrate
that that there are numerous micropores in diameters around
6–8 nm in the hollow fibers, whereas there are limited micropores
in the solid fibers. It was reported that PMMA decomposed without
carbon residue during the carbonization of PAN [41]. It is supposed
that the small molecule gaseous products from the decomposition
of PMMA penetrate through the fiber wall, leaving the extra
micropores in the hollow carbon fibers. The pore volume of the
hollow fibers is detected to be 0.18 cm3/g, and that of the solid
fibers is only 0.09 cm3/g. As the diameters of the hollows are in
the range of 30–60 nm (Fig. 2a), which are much larger than the
detected micropore sizes of 6–8 nm (Fig. 3b), it is noted that the
volume of the hollows is not involved in the detected pore volume
for the hollow fibers. The specific surface area of the hollow fibers
calculated by the BET method is 443 m2/g, whereas that of the solid
ones is only 215 m2/g.
2. Experimental
PAN (Sigma–Aldrich) and PMMA (Aladdin) were respectively dissolved in N,
N-dimethylformamide (DMF, Aldrich) by stirring at 80 °C for 24 h in concentrations
of 8 and 16 wt.%, which were used as the outer and inner solutions for coaxial elec-
trospinning. The outer and inner diameters of the nozzle were 1.2 and 0.4 mm,
respectively. The work distance was 15 cm, and the work voltage was 16.8 kV.
The feed rates of the outer and inner solutions were 1.0 and 0.5 mL/h, respectively.
The obtained fibrous webs were collected by a Ni foil and were removed every 10 h.
The fibrous webs were stabilized by a thermal treatment in an air atmosphere
oven at 250 °C for 1 h with a heating rate of 5 °C/min in order to convert PAN from
thermo plastic to non-plastic compound, which is critical to obtain dimensional sta-
bility and, hence, a fibrous morphology [20]. The stabilized fibrous webs were sub-
sequently carbonized at 900 °C for 1 h with a heating of 5 °C/min in a nitrogen
atmosphere furnace, forming MhMpCFs with the elimination of PMMA. A quartz
crucible was used as the holder for the stabilization and carbonization. The carbon-
ized fibrous webs were ground by mortar and pestle. Therefore, the fibrous webs
were broken, forming carbon fibers with open terminals, which are readily for
immersion of S. Sublimed S (Aladdin, 99%) was incorporated to the carbon nanofi-
bers by heating the mixture of S and the carbon fibers at 155 °C for 24 h in a sealed
container, which was assembled in an Ar-filled glove box (Labstar, Braun, Ger-
many). The obtained S/carbon fiber composite was taken out after the container
was cooled to room temperature. A S to carbon fiber weight ratio of 3:2 was used.
Schematic illustration of the synthesis processes of the S/MhMpCFs composite is
shown in Scheme 1.
The morphologies of the as-spun webs, ground carbonized fibers and S/carbon
fiber composite were observed by scanning electron microscopy (SEM, S-4800,
Hitachi, Japan). The crystal structure of the samples was analyzed by X-ray diffrac-
tion (XRD, X’Pert PRO, PANalytical, Nederland). The pore volume, pore size distribu-
tion and specific surface area of the carbon fibers and the S/carbon fiber composites
were analyzed by a N2 sorption analyzer (NOVA 1000e, USA) at 77 K combining the
Brunauer Emmett Teller (BET) method. Raman measurement of the carbon fibers
was performed on a confocal Mirco-Raman spectrometer (inVia-Reflex, Renishaw
plc) with a 514 nm wavelength incident laser light. The S loaded in the S/carbon
fiber composite was characterized by thermogravimetric measurement (TG,
NETZSCH QMS 403C, Germany).
Electrochemical testing of the S/MhMpCFs composite was carried out by using
coin cells of CR2025 type with Li foil as anode and counter electrode. The cathode
was prepared by coating the slurry of the S/MhMpCFs composite, acetylene black
(AB) conductive additive and polyvinylidene fluoride (PVDF) binder in a weight
ratio of 8:1:1 in N-methyl-2-pyrrolidone (NMP) on an aluminum foil and followed
by a heating at 60 °C in vacuum for 24 h. The loading of the active material of S/
MhMpCFs composite in the electrode was ca. 1.8 mg/cm2. Cells were assembled
XRD analysis of the carbonized hollow and solid fibers shows
that there is only a hump at diffraction angle 2h around 23° in their
patterns, corresponding to the general disordered stacking carbon.
The hollow and solid fibers show no intrinsic difference in their
XRD patterns. The pattern of the porous fibers is representatively

222
Y. Wu et al. / Journal of Alloys and Compounds 608 (2014) 220–228
Decomposition
of PMMA
Porous
Sulfur
Electrospinning
Carbonization
Sulfur
155 °C
PMMA
PAN
900 °C
PMMA
Carbon
PAN
Decomposition
of PMMA
(a)
(b)
(c)
(d)
Scheme 1. Schematic illustration of the synthesis processes of the S/MhMpCFs composite: (a) electrostatic spinning process; (b) as-spun fiber of PAN/PAMMA; (c) carbonized
mesohollow and microporous nanofiber; (d) S incorporated hierarchically porous carbon nanofiber composite.
shown in Fig. 4. Further Raman spectra measurement of the car-
bonized fibers showed that there were typical D band around
1360 cmÀ1 for disordered carbon and G band around 1590 cmÀ1
for graphene (the spectra were omitted here), confirming the for-
mation of carbon fibers.
ignored (Fig. 3a). Both the decreased specific surface area and the
residual pore volume of the S/SCFs composite indicate that S also
fills mostly in the micropores of the solid carbon fibers. However,
as the limited total volume of the micropores, considerable amount
of S incorporated in between the carbon fibers and bundled the
carbon fibers, which also led to the low specific surface area of
the composite.
The SEM morphologies of the hollow and solid fibers incorpo-
rated with S are shown in Fig. 2c and d, respectively. As seen from
Fig. 2c, considerable hollow carbon fibers were almost fully filled
with S after the thermal treatment. Nevertheless, hollows still exist
in many fibers. However, the diameters of the residual hollows are
evidently reduced compared with the original ones. The inset in
Fig. 2c is a magnified image of the rectangular area, showing
clearly the dense S-incorporated fibers and the residual hollow in
some other fibers. Due to the high capillary force of the molten S
in the mesohollows, considerable amount of S fills in the hollows.
In addition, some of the S-incorporated fibers shows undulating
surface, which are much clearly seen in the inset in Fig. 2c. Such
imbricate coating is suggested to be the S coating. Meanwhile, con-
siderable fibers show relatively smooth surface. However, the rel-
atively smooth surface does not mean there is no S coating. It is
also probably that the coating is much even. Moreover, congre-
gated S in between the carbon fibers, as marked by the solid arrows
in Fig. 2c, is also found in the S/MhMpCFs composite. There is
residual S incorporated outside the carbon fibers, which means
that S is sufficient for the hollows and pores. Hollows could not
be fully filled with S in the case that the loaded S was sufficient
was also found in a S/tubular polypyrrole system [42]. For the solid
carbon fibers permeated with S (S/SCFs), SEM observation shows
that there is more amount of congregated S in between the carbon
fibers. As seen in Fig. 2d, some of the carbon fibers are bundled
together by the congregated S.
The specific surface area of the S/MhMpCFs composite calcu-
lated by the BET method is only 6.0 m2/g, which is decreased by
nearly two orders compared with the pristine hollow and porous
fibers (443 m2/g). The calculated residual pore volume of the S/
MhMpCFs composite is only 0.02 cm3/g, which is also evidently
lower than that of 0.18 cm3/g for the pristine hollow fibers. The
results demonstrate that most of the micropores are filled with S.
The capillary force of the molten S in the micropores is extremely
high. The specific surface area of the S/SCFs composite is 1.1 m2/g,
which is also greatly reduced compared with that of the pristine
solid fibers (215 m2/g). However, the value is much smaller than
that of the S/MhMpCFs composite, as there are still residual hol-
lows in the S-incorporated hollow carbon fibers. In addition, the
residual pore volume in the S/SCFs composite can almost be
Fig. 5 is the TG curves of the S/MhMpCFs and S/SCFs composites
as well as the individual S. As seen from Fig. 5, the content of S
incorporated in either the S/MhMpCFs composite or S/SCFs com-
posite is almost 60 wt.%, which is identical to the amount set in
their original mixtures. However, further seen from Fig. 5, though
the visible weight loss of the individual S and the S in the
composites takes place at an almost same temperature of ca.
200 °C, the sublimation process of the S in the S/MhMpCFs com-
posite is very different from either of the individual S or the S in
the S/SCFs composite. The sublimation temperature of the S in
the S/SCFs composite is only slightly higher than that of the indi-
vidual S, and the sublimation completes at 328 °C, which is almost
the same as that of the individual S. However, the S in the S/
MhMpCFs composite shows a much lower sublimation rate, and
the sublimation completes at a much higher temperature of
469 °C. The much difficulty of the sublimation of the S in the S/
MhMpCFs composite is due to the strong adsorption force between
the S confined in the mesohollows/micropores and the carbon
matrix.
Based on the density of sulfur (2.07 g/cm3) and the pore volume
of the hollow fibers (0.18 cm3/g), it can be estimated that there is
ca. 25 wt.% of the incorporated S able to immerse in the micropores
of the hollow carbon fibers. Partial sulfur is filled in the hollow
cores and the other is coated on the fiber surface and exists as con-
glomerations in between the carbon fibers. Further seen from
Fig. 5, the TG curves of the whole sublimation processes of the
S/MhMpCFs composite and the individual S are almost fully
separate. There is no evidently overlapped part, even at the begin-
ning of the sublimation. This phenomenon suggests that the S
coated on the outer surface of the hollow fibers also shows lower
sublimation rate than the individual S. The strong absorption
derived from the porous structure is also effective for the surface
coated S. For the solid carbon fibers, as the much less pore volume
and there is no hollow in the fibers, the incorporated S exists
mostly as conglomerations in between the carbon fibers (Fig. 2d)
and is coated on the surface of the fibers in addition to the limited
trapping in the micropores. That is also the reason for the similar
sublimation characteristics with the individual S. The slightly

Y. Wu et al. / Journal of Alloys and Compounds 608 (2014) 220–228
223
capacity retention of only 47%. Activation in initial several cycles
was also found in S/acetylene black cathodes with and without
CNF (carbon nanofiber) addition using PEO (polyethylene oxide)
as binder [44]. Activation up to 20 cycles was also reported in a
S-polypyrrole/graphene multi-composite cathode [23]. Activation
was considered due to the slow penetration and transportation
of solvated electrolyte and lithium ions in the electrodes. In the
present study, the incomplete utilization of sulfur in the
S/MhMpCFs composite cathode in the initial several cycles is also
due to that the electrolyte cannot fully penetrate into the microp-
ores and mesohollows, and contact sulfur effectively. However, the
condition was solved after several cycles.
The superior cycling performance of the S/MhMpCFs composite
is attributed to the low dissolution of the soluble intermediates of
polysulfides generated from the S encapsulated in the micropores
and meshollows in the electrolyte. The mobility of the electrolyte
in the mesohollows and the micropores is low, therefore, the dis-
solved polysulfides are readily to reach a high concentration in
the electrolyte, which reduce the dissolution rate and lead to a high
utilization of the active materials. Moreover, the residual hollows
provide more channels for the transportations of electrolyte and
lithium ions, which also favor the utilization of the sulfur either
incorporated in the hollows or trapped in the micropores. Whereas
for the S/SCFs composite, the loaded S exists mostly on the surface
of the fibers and congregates in between the fibers, which exposes
in the electrolyte without defense. The dissolution of the soluble
intermediate polysulfides during cycling is severe, resulting severe
mass loss of the active materials. A capacity retention of 92% for
20 cycles at 0.1 C was reported for carbon nanofibers with single
microporous structure prepared from carbonization of PAN and
incorporated with 42 wt.% S [40]. The present S/MhMpCFs compos-
ite with 60 wt.% S provides a capacity retention of 88% for 70 cycles
at 0.1 C. The average retention for per cycle is higher than that
reported in the literature. The capacity and capacity retention of
the present S/MhMpCFs composite are also higher than those of
the composite of non-porous tubular polypyrrole (T-PPy) fibers
incorporated with 30 wt.% S [42]. It demonstrates that the combi-
nation of mesohollows and micropores in the carbon fibers is much
effective in improving the electrochemical performance of the S
cathode compared with either single microporous structure or sin-
gle hollow structure.
Fig. 1. SEM image of the as-spun webs of PAN/PMMA.
higher onset sublimation temperature of the S/SCFs composite
than that of the individual S as shown in Fig. 5 is attributed to
the few amount of S trapped in the micropores of the solid carbon
fibers. In addition, the TG curves of the S/MhMpCFs and S/SCFs
composites are almost fully separate without visible overlap,
which also indicates that the congregated S in between the carbon
fibers in the former is much less that in the latter.
XRD patterns of the S/MhMpCFs and S/SCFs composites are also
shown in Fig. 4. For comparison, the pattern of the pristine S is also
shown. All the diffraction peaks in the composites correspond to
orthorhombic S (JCPDS#08-0247). There is only S, no other impu-
rity, detected. It was reported that the S confined in micropores
in nanosize showed no evident diffraction peaks due to their amor-
phous or nanocrystalline feature [29,43]. The presence of the dif-
fraction peaks of crystalline S in the S/MhMpCFs composite is
attributed to the S filled in the hollow cores and the S incorporated
outside the carbon fibers, including both coated on the surface of
the fibers and congregated in between the fibers. In addition, it is
certain that the S in the S/SCFs composite shows crystal structure
as it exists mostly on the outer surface of the carbon fibers and
as conglomerations in between the carbon fibers. The main crystal-
line facet indexes of S are marked in the patterns of the composites
and pristine S. The lattice parameters and cell volume of the S in
both composites and the pristine S were calculated from the XRD
data and are listed in Table 1. It is seen that the values of the S
in the two composite are very close, whereas they are slightly
lower than those of the pristine S. It seems that with the incorpo-
ration with the carbon fibers, the lattice parameter of the S is
slightly reduced.
Fig. 7a and b depicts the discharge/charge profiles of selected
cycles of the S/MhMpCFs and S/SCFs composites, respectively. As
a general look from Fig. 7a and b, there are two plateaus in the dis-
charge curves for both composites, indicating a two-step reaction.
Generally, the first discharge plateau in the comparatively higher
potential range corresponds to the conversion of S (in the first
cycle) or high-order polysulfides (e.g. Li2S8, in the subsequent
cycles) to low-order polysulfides (Li2Sn, 4 6 n < 8). The second
discharge plateau in the comparatively lower potential range
corresponds to the conversion of the low-order polysulfides to
lithium sulfides (LiS2/Li2S2) [41,43,45].
3.2. Electrochemical properties
As seen from Fig. 7a, the two plateaus in the first discharge pro-
cess of the S/MhMpCFs composite are all considerably lower than
those in the subsequent cycles up to 70 cycles, which is attributed
to the large polarization caused by the insufficient penetration of
electrolyte in the composite in the first cycle. Both plateaus rise
to higher potential ranges in the second discharge, indicating an
improved lithium reaction kinetics, due to the improved penetra-
tion of electrolyte in the composite and also probably the change
in structure of the active materials, such as from S8 to high-order
polysulfides (e.g. Li2S8) [43]. Further seen from Fig. 7a, the dis-
charge capacities generated from the first plateau (in the compar-
atively higher potential range) increases from ca. 230 mA h/g at the
first cycle to the maximum value of ca. 270 mA h/g at the 6th cycle
due to the activation. Thereafter, the capacity decreases with
Fig. 6 shows the cycle performance of the S/MhMpCFs as well as
the S/SCFs composite cathodes. As seen from Fig. 6, the first dis-
charge capacity of the S/MhMpCFs composite is only 640 mA h/g,
which is much lower than 940 mA h/g of the S/SCFs composite.
However, the capacity of the S/MhMpCFs composite increases con-
tinuously in the subsequent cycles up to 6 cycles, reaching a max-
imum value of 815 mA h/g. Thereafter, the S/MhMpCFs composite
shows considerable stable cycling stability, and the capacity
remains 715 mA h/g after 70 cycles, corresponding to a capacity
retention of 88% with respect to the maximum capacity. Mean-
while, the S/SCFs composite shows continuous drop in the capacity
with cycling, and the capacity drops to 730 mA h/g at the 6th cycle.
The capacity is only 428 mA h/g after 70 cycles, corresponding to a

224
Y. Wu et al. / Journal of Alloys and Compounds 608 (2014) 220–228
Fig. 2. SEM images of the carbonized fibers from PAN/PMMA (a), carbonized fibers from single PAN (b), S/MhMpCFs (c) and S/SCFs (d) composites.
140
0.06
(a)
(b)
120
100
80
60
40
20
0
0.05
0.04
0.03
0.02
0.01
0.00
MhMpCFs
SCFs
MhMpCFs
SCFs
S/MhMpCFs
S/SCFs
S/MhMpCFs
S/SCFs
0.0
0.2
0.4
0.6
0.8
1.0
4
6
8
10
12
14
16
Relative Pressure (P/P0)
Pore diameter (nm)
Fig. 3. N2 adsorption isotherms (a) and pore size distribution (b) of the hollow and solid carbonized fibers and their composites with S.
(026)
(222)
(040)
(040)
S/MhMpCFs
S/SCFs
cycling and to a value of ca. 220 mA h/g at the 70th cycle. Whereas
for the S/SCFs composite, the capacity generated from the first dis-
charge plateau decreases in a large extent from ca. 300 to
150 mA h/g from the first cycle to the 70th cycle. This result indi-
cates that the reversibility of the conversion reaction of S8/high-
order polysulfides to low-order Li2Sn (4 6 n < 8) of the S/MhMpCFs
composite is much higher than that of the S/SCFs composite, con-
firming a less loss of the active material during cycling for the for-
mer. In addition, the length of the second plateau (in the
comparatively lower potential range) of the S/MhMpCFs composite
also extends to a maximum value, and the plateau seems to be the
flattest at the 6th cycle. It is noted that each plateau generates the
maximum capacity at the 6th cycles, and results in totally the
maximum capacity of the composite. Thereafter, either the length
or the plateau potentials decrease continuously with the cycling,
corresponding to a decreasing capacity and an increasing polariza-
tion. Moreover, the discharge plateau in the low potential range is
(222)
(222)
(026)
(026)
(040)
S
porous carbon fibers
S: JCPDS 08-0247
10
20
30
40
50
2θ (°)
60
70
80
90
Fig. 4. XRD patterns of the hollow carbon fibers, the S/MhMpCFs and S/SCFs
composites as well as pristine S.

Y. Wu et al. / Journal of Alloys and Compounds 608 (2014) 220–228
225
always much longer than the one in the high potential range. The
conversion reaction from the low-order polysulfides to lithium
sulfides represents the major discharge capacity of composite, as
the common Li–S batteries.
individual sulfur
S/SCFs
100
80
60
40
20
0
S/MhMpCFs
The charge process is associated with the reversible conversion
of Li2S/Li2S2 to low-order polysulfides and then to high-order
polysulfides. The two-step conversion feature is not evident in
the initial several cycles for the S/MhMpCFs composite. The high
overlap of the plateaus for the different conversion reactions is
probably due to the high overvoltage for the conversion of Li2S/
Li2S2 to Li polysulfides [46]. However, two-plateau feature
becomes visible after initial several cycles and tends to be much
evident with the further cycling. In addition, lithium extraction
from the active material on the outer surface of the fibers is much
easier than from that confined in the micropores and mesohol-
lows of the fibers. With the preferred dissolution of the polysul-
fides incorporated on the outer surface of the fibers, the amount
of the active material incorporated on the surface of the fibers
reduced. The much visibly raised plateau in the comparatively
higher potential range as seen in Fig. 7a, which corresponds to
the conversion reaction of low-order polysulfides to high-order
polysulfides, is associated with the much difficulty for the extrac-
tion of lithium from the polysulfides confined in the micropores
and mesohollows, i.e., associated with an increased polarization.
Whereas the plateau corresponding to the conversion reaction
of Li2S/Li2S2 to low-order polysulfides shows only very limited
rise in the potential during cycling, corresponding to a less dete-
rioration in polarization. In addition, the conversion of the low-
order polysulfides to high-order polysulfides generates the major
charge capacity as seen also from Fig. 7a.
For the S/SCFs composite, as seen from Fig. 7b, there are two
plateaus with limited flat range in the first discharge. The dis-
charge plateau corresponding to the conversion reaction of
S8/high-order polysulfides to low-order Li2Sn (4 6 n < 8) rises in a
very early stage of the reaction upon cycling, i.e., rise occurs only
in a very high potential range, thereafter the plateau drops and
to larger extent with cycling, indicating an increasing polarization.
Whereas, the plateau corresponding to the conversion reaction of
low-order polysulfides to Li2S/Li2S2 shows evidently raised flat fea-
ture in the subsequent cycles, and the plateau potential is almost
not changed with cycling, indicating a facile lithium insertion.
However, the length of the plateau is evidently shortened with
cycling, indicating a rapid loss of the active material and thus a
rapid decrease in capacity. For the charge process, there is only
one flat plateau observed, and the plateaus are highly overlapped,
indicating that the lithium extraction takes place in an almost
same potential upon cycling.
0
100
200
300
400
500
600
Temperature (oC)
Fig. 5. TG curves of the S/MhMpCFs and S/SCFs composites as well as individual S.
Table 1
Lattice parameter and cell volume of the S in the composites and pristine S.
Samples
Lattice parameters (Å)
Cell volume (Å3)
a
b
c
S/MpMhCFs
S/SCFs
Pristine S
10.46
10.47
10.49
12.86
12.87
12.96
24.47
24.53
24.65
3291.40
3306.13
3352.74
1000
800
600
400
200
0
S/MhMpCFs
S/SCFs
0
10
20
30
40
50
60
70
Cycle number
Fig. 6. Cycling performance of the S/MhMpCFs and S/SCFs composites.
3.2
3.0
2.8
2.6
2.4
2.2
2.0
1.8
1.6
1.4
3.2
3.0
2.8
2.6
2.4
2.2
2.0
1.8
1.6
1.4
(b)
(a)
1
1
2
2
6
6
25
50
70
25
50
70
0
200
400
600
800
1000
0
200
400
600
800
1000
Capacity (mAh/g)
Capacity (mAh/g)
Fig. 7. Voltage profiles of selected cycles of the S/MhMpCFs (a) and S/SCFs (b) composites.

226
Y. Wu et al. / Journal of Alloys and Compounds 608 (2014) 220–228
The polysulfides in the S/MhMpCF composite suffer from less
it is noteworthy that the CV analysis is generally effective in
characterizing the step-feature of the redox reactions of the elec-
trode. Two overlapped anodic peaks for S cathode were also
reported in previous literature [51]. Moreover, it seems that there
is only one anodic peak centered at ca. 2.40 V in the anodic pro-
cess of the S/SCFs composite, which is also well consistent with
the one flat charge plateau in the charge curve (Fig. 7b). The
slightly lower anodic peak potential of the S/SCFs composite than
that of the S/MhMpCFs composite (2.39–2.40 V) is also well con-
sistent with their charge plateaus. The one anodic peak feature is
supposed due to the facile conversion of Li2S/Li2S2 to high-order
polysulfides, which consists of multistep oxidation reactions.
One anodic peak feature is also commonly reported in other S
cathodes [43,52]. In addition, there is no anodic peak centered
at ca. 2.8 V (vs. Li+/Li) found in the present study for the forma-
tion of S8 as reported in a S loaded hierarchically pore-structured
carbon pillars cathode [29], indicating that there is likely no S8
formed in the present final lithium extraction product. In the sec-
ond cathodic process, the cathodic process moves to higher
potential ranges for both composites, especially for the peak
located in the relatively higher potential, which is mainly attrib-
uted to the change of the active material from S8 to high-order
polysulfides. The phenomenon is also in good agreement with
the raised discharge plateaus in the discharge curves.
dissolution in the electrolyte due to the confinement, which
improves the utilization of the active materials. In addition, the
strong interaction between the active materials encapsulated in
the micropores/mesohollows and the carbon matrix are all benefi-
cial in improving the electron contact of the active materials,
which are also favorable to the electrochemical property of the
S/MhMpCFs composite. The residual hollows in the S-incorporated
hollow fibers enhance the penetration of the electrolyte to the
active materials incorporated in the inner wall of the fibers, favor-
ing the mobility of lithium ions for the electrochemical reaction,
and provide rooms for alleviating the volume change of the active
materials during cycling. As a result, the S/MhMpCFs composite
shows favorable capacity and cycling stability as cathode material
for Li–S batteries. Further more, it is considered that there is also
room for adjusting the micropore size and pore volume of the por-
ous carbon fibers by adjusting the fraction of PAN and PMMA used
for the electrospinning to improve the overall electrochemical
property of S/MhMpCFs composite in the future study. Increasing
the operating voltage of the S cathode is important for practical
applications. Though the operating voltage for S cathode is intrin-
sically low, exploring approaches to achieve comparatively high
operating voltage is still possible. Introducing high conductive
matrix to improve the electrical contact of the S active materials
is demonstrated effective in improving the discharge potential
(vs. Li+/Li) [47,48]. Application of suitable electrolyte to reduce
polarization is also important to achieve comparatively high dis-
charge potential [49].
20
Fig. 8a and b shows the CV curves of the S/MhMpCFs and
S/SCFs composites of the first three cycles at a scanning rate of
0.1 mV/s. Both composites show two cathodic peaks correspond-
ing to the typical two-step conversion reactions [18,50]. The one
centered around 2.3 V corresponds to the transformation of S8 to
Li2Sn (4 6 n < 8), and the other centered around 2.0 V corresponds
the further reduction of Li2Sn (4 6 n < 8) to Li2S/Li2S2. In addition,
the cathodic peak positions of the S/MhMpCFs composite are
slightly lower than those of the S/SCFs composite, as shown in
Fig. 8, indicating slightly lower reaction kinetics for the former.
There are two partially overlapped anodic peaks in the
S/MhMpCFs composite. The main one is centered at ca. 2.42 V
(vs. Li+/Li), and the minor one is centered at a slightly lower value
of 2.39 V, which correspond to the conversions of Li2S/Li2S2 to
low-order polysulfides and the low-order polysulfides to high-
order polysulfides, respectively. The partially overlapped anodic
peak feature is also consistent with the phenomenon that the
two charge plateaus show very close potentials in the charge
curves, especially for the initial several cycles (Fig. 7a). However,
S/SCFs
15
10
5
S/MhMpCFs
R s
CPE 1
R f
C PE2
R ct
W 0
0
0
5
10
15
20
25
30
Z' (ohm)
Fig. 9. Electrochemical impedance spectra of the S/MhMpCFs and S/SCFs compos-
ites. The inset is the equivalent circuit. The solid symbols denote the experimental
data, and the lines represent the fitting results from the calculation based on the
equivalent circuit.
2.0
2.0
2.40
2.42
(b)
1st cycle
2nd cycle
3rd cycle
(a)
1st cycle
2nd cycle
3rd cycle
1.6
1.2
1.6
1.2
2.39
0.8
0.8
S/solid fibers
0.4
0.4
S/porous fibers
2.02
0.0
0.0
-0.4
-0.8
-0.4
-0.8
2.04
2.02
2.31
2.30
2.24
2.26
2.04
1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2
1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2
Potential vs. Li+/Li (V)
Potential vs. Li+/Li (V)
Fig. 8. CV curves of the S/MhMpCFs (a) and S/SCFs (b) composites of the first three cycles at a scanning rate of 0.1 mV/s.

Y. Wu et al. / Journal of Alloys and Compounds 608 (2014) 220–228
227
Fig. 9 illustrates the electrochemical impedance analyses of the
References
S/MhMpCFs and S/SCFs composites. The solid symbols denote the
experimental data, and the lines represent the fitting results from
the calculation based on the equivalent circuit inserted in Fig. 9.
The experimental data show that the electrochemical impedance
spectra of the two composites all consist of two depressed semicir-
cles in the high and medium frequency ranges followed by a short
line in the low frequency range. The intercept of the real axis at the
high frequency is ascribed to the ohmic resistance resulting from
electrolyte resistance and cell components [53,54], which is
denoted as Rs. The semicircle from high to medium frequency rep-
resents the resistance from the SEI film on the surface of the cath-
ode [54], which is denoted as Rf. The second semicircle at medium
frequency relates to the charge transfer resistance [54], Rct. The
short line in the low frequency region is generally named as War-
burg resistance, denoted as W0, which is attributed to the diffusion
resistance of lithium ions within the S cathode. CPE1 and CPE2 in
the equivalent circuit represent the capacitive elements [53]. As
seen from Fig. 9, the radiuses of the two semicircles of the
S/MhMpCFs composite are all smaller than those of the S/SCFs
composite. The calculated values of Rf and Rct of the S/MhMpCFs
[1] P.G. Bruce, L.J. Hardwick, K. Abraham, MRS Bull. 36 (2011) 506–512.
[2] M.Y. Chu, Rechargeable positive electrodes, Google Patents, 1998.
[3] D. Marmorstein, T. Yu, K. Striebel, F. McLarnon, J. Hou, E. Cairns, J. Power
Sources 89 (2000) 219–226.
[4] H. Yamin, E. Peled, J. Power Sources 9 (1983) 281–287.
[5] L. Yin, J. Wang, F. Lin, J. Yang, Y. Nuli, Energy Environ. Sci. 5 (2012) 6966–6972.
[6] Y.V. Mikhaylik, J.R. Akridge, J. Electrochem. Soc. 151 (2004) A1969–A1976.
[7] S.E. Cheon, K.S. Ko, J.H. Cho, S.W. Kim, E.Y. Chin, H.T. Kim, J. Electrochem. Soc.
150 (2003) A796–A799.
[8] S.E. Cheon, K.S. Ko, J.H. Cho, S.W. Kim, E.Y. Chin, H.T. Kim, J. Electrochem. Soc.
150 (2003) A800–A805.
[9] F. Wu, J. Chen, L. Li, T. Zhao, R. Chen, J. Phys. Chem. C 115 (2011) 24411–24417.
[10] L. Xiao, Y. Cao, J. Xiao, B. Schwenzer, M.H. Engelhard, L.V. Saraf, Z. Nie, G.J.
Exarhos, J. Liu, Adv. Mater. 24 (2012) 1176–1181.
[11] X. Liang, Z. Wen, Y. Liu, X. Wang, H. Zhang, M.W. Huang, Solid State Ionics 192
(2011) 347–350.
[12] F. Wu, J. Chen, R. Chen, S. Wu, L. Li, S. Chen, T. Zhao, J. Phys. Chem. C 115 (2011)
6057–6063.
[13] F. Wu, S. Wu, R. Chen, J. Chen, S. Chen, Electrochem. Solid-State Lett. 13 (2010)
A29–A31.
[14] J. Guo, Y. Xu, C. Wang, Nano Lett. 11 (2011) 4288–4294.
[15] W. Ahn, K.B. Kim, K.N. Jung, K.H. Shin, C.S. Jin, J. Power Sources 202 (2012)
394–399.
[16] L. Wang, Y. Zhao, M.L. Thomas, H.R. Byon, Adv. Funct. Mater. (2013), http://
dx.doi.org/10.1002/adfm.201302915.
[17] G. He, S. Evers, X. Liang, M. Cuisinier, A. Garsuch, L.F. Nazar, ACS Nano 7 (2013)
10920–10930.
[18] G. Ma, Z. Wen, J. Jin, Y. Lu, K. Rui, X. Wu, M. Wu, J. Zhang, J. Power Sources 254
(2014) 353–359.
[19] D. Li, F. Han, S. Wang, F. Cheng, Q. Sun, W.C. Li, ACS Appl. Mater. Interfaces 5
(2013) 2208–2213.
[20] C. Zhang, H.B. Wu, C. Yuan, Z. Guo, X.W.D. Lou, Angew. Chem. Int. Ed. 124
(2012) 9730–9733.
[21] X. Wang, Z. Wang, L. Chen, J. Power Sources 242 (2013) 65–69.
[22] W. Zhou, H. Chen, Y. Yu, D. Wang, Z. Cui, F.J. DiSalvo, H.D. Abruña, ACS Nano 7
(2013) 8801–8808.
composite are 9.6 and 8.1
lower than 17.1 and 69.0
X
X
, respectively, which are significantly
of the S/SCFs composite. Especially
for the charge transfer resistance, the value of the S/MhMpCFs
composite is nearly one order smaller than that of the S/SCFs
composite. The much smaller charge transfer resistance of the
S/MhMpCFs composite indicates a much higher electronic conduc-
tivity and a much better reaction kinetics of the composite. The
result further demonstrates that the mesohollow and microporous
structure of the carbon fibers provides better electronic connection
for the incorporated S than the solid structured carbon fibers.
[23] W. Wang, G.C. Li, Q. Wang, G.R. Li, S.H. Ye, X.P. Gao, J. Electrochem. Soc. 160
(2013) A805–A810.
[24] H. de Vries, S. Jeong, M. Joost, M. Keller, S. Passerini, J. Electrochem. Soc. (2013).
379–379.
[25] S. Wei, H. Zhang, Y. Huang, W. Wang, Y. Xia, Z. Yu, Energy Environ. Sci. 4 (2011)
736–740.
4. Conclusions
[26] G. He, X. Ji, L. Nazar, Energy Environ. Sci. 4 (2011) 2878–2883.
[27] C. Liang, N.J. Dudney, J.Y. Howe, Chem. Mater. 21 (2009) 4724–4730.
[28] S.R. Chen, Y.P. Zhai, G.L. Xu, Y.X. Jiang, D.Y. Zhao, J.T. Li, L. Huang, S.G. Sun,
Electrochim. Acta 56 (2011) 9549–9555.
[29] X. Wang, X. Fang, X. Guo, Z. Wang, L. Chen, Electrochim. Acta 97 (2013) 238–
243.
[30] N. Jayaprakash, J. Shen, S.S. Moganty, A. Corona, L.A. Archer, Angew. Chem. Int.
Ed. 123 (2011) 6026–6030.
[31] R. Demir-Cakan, M. Morcrette, F. Nouar, C. Davoisne, T. Devic, D. Gonbeau, R.
Dominko, C. Serre, G. Férey, J.M. Tarascon, J. Am. Chem. Soc. 133 (2011)
16154–16160.
[32] W. Bao, Z. Zhang, Y. Qu, C. Zhou, X. Wang, J. Li, J. Alloys Comp. 582 (2014) 334–
340.
[33] S. Li, C. Chen, K. Fu, L. Xue, C. Zhao, S. Zhang, Y. Hu, L. Zhou, X. Zhang, Solid State
Ionics 254 (2014) 17–26.
[34] X. Xia, X. Wang, H. Zhou, X. Niu, L. Xue, X. Zhang, Q. Wei, Electrochim. Acta 121
(2014) 345–351.
[35] Y. Wu, P. Zhu, M.V. Reddy, B.V.R. Chowdari, S. Ramakrishna, ACS Appl. Mater.
Interfaces 6 (2014) 951–1958.
[36] P. Zhu, Y. Wu, M.V. Reddy, A. Sreekumaran Nair, B.V.R. Chowdari, S.
Ramakrishna, RSC Adv. 2 (2012) 531–537.
[37] Y. Wu, R. Balakrishna, M.V. Reddy, A.S. Nair, B.V.R. Chowdari, S. Ramakrishna, J.
Alloys Comp. 517 (2012) 69–74.
[38] C.T. Cherian, J. Sundaramurthy, M.V. Reddy, P.S. Kumar, K. Mani, D. Pliszka, C.H.
Sow, S. Ramakrishna, B.V.R. Chowdari, ACS Appl. Mater. Interfaces 5 (2013)
9957–9963.
[39] Y. Wu, P. Zhu, X. Zhao, M.V. Reddy, S. Peng, B.V.R. Chowdaric, S. Ramakrishna, J.
Mater. Chem. A 1 (2013) 852–859.
Hierarchically porous carbon fibers in diameters of 200–300 nm
with mesohollows of 30–60 nm and micropores of several nano-
meters have been successfully synthesized by coaxial electrospin-
ning PAN and PMMA as the outer and inner spinning solutions,
respectively, followed by a carbonization. PMMA decomposed in
the carbonization process and induced the mesohollows and
micropores. S fills fully in the micropores and fills mostly in the
mesohollows via the thermal treatment, forming hierarchical
structured-S/carbon fiber composite with 60 wt.% S. This sulfur/
porous carbon fiber composite shows a maximum capacity of
815 mA h/g at 0.1 C after several activation cycles as a cathode
material for Li–S batteries, and a capacity retention of 88% is
obtained after 70 cycles with respect to the maximum capacity.
The cycling stability is much better than that of the sulfur/solid
carbon fiber composite cathode, the solid fibers of which were pre-
pared from individual PAN. The micropores, which encapsulate
effectively the S active material and alleviate the dissolution of sol-
uble polysulfides in the electrolyte, and the mesohollows, which
allow a sufficient penetration of electrolyte for the mobility of lith-
ium ions, play synergetic roles in improving the electrochemical
property of the S/porous carbon fiber composite.
[40] L. Ji, M. Rao, S. Aloni, L. Wang, E.J. Cairns, Y. Zhang, Energy Environ. Sci. 4
(2011) 5053–5059.
[41] C. Kim, Y.I. Jeong, B.T.N. Ngoc, K.S. Yang, M. Kojima, Y.A. Kim, M. Endo, J.W. Lee,
Small 3 (2007) 91–95.
Acknowledgments
[42] X. Liang, Y. Liu, Z. Wen, L. Huang, X. Wang, H. Zhang, J. Power Sources 196
(2011) 6951–6955.
This work was supported by National Nature Science Founda-
tion for Distinguished Youth Scholars of China (No. 51025102),
National Nature Science Foundation of China (No. 51371158),
Program for Innovative Research Team in University of Ministry
of Education of China (IRT13037) and Key Science and Technology
Innovation Team of Zhejiang Province (2010R50013).
[43] X. Cui, Z. Shan, L. Cui, J. Tian, Electrochim. Acta 105 (2013) 23–30.
[44] Y.J. Choi, K.W. Kim, H.J. Ahn, J.H. Ahn, J. Alloys Comp. 449 (2008) 313–316.
[45] Y. Yang, G. Zheng, Y. Cui, Chem. Soc. Rev. 42 (2013) 3018–3032.
[46] Y. Cao, X. Li, I.A. Aksay, J. Lemmon, Z. Nie, Z. Yang, J. Liu, Phys. Chem. Chem.
Phys. 13 (2011) 7660–7665.
[47] W. Bao, Z. Zhang, C. Zhou, Y. Lai, J. Li, J. Power Sources 248 (2014) 570–
576.

228
Y. Wu et al. / Journal of Alloys and Compounds 608 (2014) 220–228
[48] S. Jiang, Z. Zhang, X. Wang, Y. Qu, Y. Lai, J. Li, RSC Adv. 3 (2013) 16318–16321.
[49] L. Suo, Y. Hu, H. Li, M. Armand, L. Chen, Nat. Commun. 4 (2013). 1481.1–9.
[50] S. Chen, X. Huang, H. Liu, B. Sun, W. Yeoh, K. Li, J. Zhang, G. Wang, Adv. Energy
Mater. (2014), http://dx.doi.org/10.1002/aenm.201301761.
[52] X. Tao, X. Chen, Y. Xia, H. Huang, Y. Gan, R. Wu, F. Chen, J. Mater. Chem. A 1
(2013) 3295–3301.
[53] N.A. Canas, K. Hirose, B. Pascucci, N. Wagner, K.A. Friedrich, R. Hiesgen,
Electrochim. Acta 97 (2013) 42–51.
[51] J. Kim, D.J. Lee, H.G. Jung, Y.K. Sun, J. Hassoun, B. Scrosati, Adv. Funct. Mater. 23
(2013) 1076–1080.
[54] J. Zheng, M. Gu, H. Chen, P. Meduri, M.H. Engelhard, J. Zhang, J. Liu, J. Xiao, J.
Mater. Chem. A 1 (2013) 8464–8470.