Carbon-coated Si as a lithium-ion battery anode material

Article abstract of DOI:10.1149/1.1518988

Carbon-coated Si has been prepared by a thermal vapor decomposition method. Its electrochemical performance has been investigated by charge/discharge tests, cyclic voltammetric experiments, differential scanning calorimetry, and ⁷Li-nuclear magnetic resonance, etc. This kind of material demonstrates good electrochemical performance as an anode material for lithium-ion batteries. The improvement in the electrochemical performance of Si is mainly attributed to the effect of carbon coating.

Full text of DOI:10.1149/1.1518988

Carbon-Coated Si as a Lithium-Ion Battery Anode Material
Masaki Yoshio, Hongyu Wang, Kenji Fukuda, Tatsuo Umeno, Nikolay Dimov and
Zempachi Ogumi
J. Electrochem. Soc. 2002, Volume 149, Issue 12, Pages A1598-A1603.
doi: 10.1149/1.1518988
Receive free email alerts when new articles cite this article - sign up in the
box at the top right corner of the article or click here
Email alerting
service
To subscribe to Journal of The Electrochemical Society go to:
http://jes.ecsdl.org/subscriptions
© 2002 ECS - The Electrochemical Society
Downloaded on 2013-02-24 to IP 137.99.31.134 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

A1598
Journal of The Electrochemical Society, 149 ͑12͒ A1598-A1603 ͑2002͒
0013-4651/2002/149͑12͒/A1598/6/$7.00 © The Electrochemical Society, Inc.
Carbon-Coated Si as a Lithium-Ion Battery Anode Material
,z
c
c
Masaki Yoshio,a, Hongyu Wang,b, Kenji Fukuda, Tatsuo Umeno,
*
*
Nikolay Dimov,a and Zempachi Ogumib,
*
aDepartment of Applied Chemistry, Saga University, Saga 840-8502, Japan
bDepartment of Energy and Hydrocarbon Chemistry, Graduate School of Engineering, Kyoto University,
Kyoto 606-8501, Japan
cMitsui Mining Company Limited, Research Center, Wakamatsu-ku, Kitakyushu 808-0021, Japan
Carbon-coated Si has been prepared by a thermal vapor decomposition method. Its electrochemical performance has been inves-
tigated by charge/discharge tests, cyclic voltammetric experiments, differential scanning calorimetry, and 7Li-nuclear magnetic
resonance, etc. This kind of material demonstrates good electrochemical performance as an anode material for lithium-ion
batteries. The improvement in the electrochemical performance of Si is mainly attributed to the effect of carbon coating.
© 2002 The Electrochemical Society. ͓DOI: 10.1149/1.1518988͔ All rights reserved.
Manuscript submitted February 25, 2002; revised manuscript received June 4, 2002. Available electronically October 29, 2002.
Graphite has been commonly used as the anode material for
zene vapor decomposed and deposited onto the surface of Si par-
ticles as a carbon coating. After a certain time of carbon coating, the
furnace was switched off and cooled down slowly. The collected
sample was softly ground by a pestle and mortar, and then fed into
the tube furnace subject to TVD treatment again. After several
cycles repeat of the above procedure, the resultant carbon-coated Si
as well as the original Si was tested as the anode material for
lithium-ion batteries. Some physical properties of the carbon-coated
Si sample are listed in Table I.
Si-based electrodes were prepared by spreading the carbon-
coated Si powder ͑or the original Si powder͒ slurry ͑90 wt %͒ and
poly͑vinylidene fluoride͒ ͑PVDF, 10 wt %͒ dissolved in 1-methyl-
2-pyrrolidinone onto a copper foil substrate. The electrodes were
then dried overnight at 105°C under vacuum and pressed between
two flat stainless steel plates.
The electrolytes of 1 M LiPF6-ethylene carbonate: dimethyl car-
bonate ͑EC:DMC͒ ͑1:2 by volume͒ and 1 M LiPF6-PC were used in
our studies. The water contents in these electrolytes were Ͻ20 ppm.
The electrochemical measurements included charge/discharge
and cyclic voltammetric ͑CV͒ tests on the Si-based electrodes. In the
charge/discharge tests, two-electrode cells were used. The two-
electrode cells generally included a lithium metal electrode and a
Si-based electrode, held apart by a separator ͑Celgard 2400͒ and a
glass fiber. For simulating the true situations in practical lithium-ion
batteries, the lithium metal electrode in the two-electrode cells was
replaced by LiMn2O4 , LiCoO2 , or LiNiO2 electrodes in some
cases. The electrodes of LiMn2O4 were received from Tohso and the
LiCoO2 and LiNiO2 electrodes were obtained from Nihon Chemical
Industry.
The charge/discharge tests included two sets of procedures. In
the first set, namely, constant capacity charge ͑Here ‘‘charge’’ cor-
responds to lithium intercalation into Si-based electrodes͒ modes,
the procedures were as follows ͑e.g., for carbon-coated Si/Li metal
cell͒: (i) The cell was charged at a constant current density of
0.5 mA/cm2 from open circuit voltage ͑OCV, ca. 3 V͒ until the
desired capacity, say 800 mAh/g, was reached; (ii) the cell was
lithium-ion batteries by virtue of its low and flat working voltage,
good cyclability in compatible electrolytes, etc. However, its theo-
retical capacity is limited to 370 mAh/g or volumetric capacity of
830 Ah/L. To increase the specific energy of lithium-ion batteries,
alternative anode materials with higher capacity are needed.
Besenhard et al. have evaluated different lithium alloys as anode
materials for lithium-ion batteries.1 Among so many matrices for
lithium alloying, Si stands out because of its high capacity to ac-
commodate lithium ͑with the maximum uptake of Li4.4Si, corre-
sponding to 4200 mAh/g Si͒. Unfortunately, there is one severe
problem with the application of Si anode, i.e., the big volume
change during lithium intercalation/deintercalation, which inevitably
causes great stress in Si lattice and thus leads to cracking and crum-
bling of the Si particle. As a result, capacity for alloying lithium
fades abruptly only after several cycles of charge/discharge.
To circumvent this problem, several research efforts have been
carried out to disperse fine Si particles within a solid, mixed con-
ducting host matrix by mechanical milling.2-5 Considerable im-
provements in cyclability have been obtained. Nevertheless, me-
chanical milling makes the specific surface area of the composite
anode materials very big. Consequently, the decomposition of the
electrolytes to build up solid electrolyte interface ͑SEI͒ films cover-
ing the surface of the composite anodes at low voltage will become
drastic, and thus give rise to large irreversible capacities at the initial
several cycles. This problem can still preclude further use of Si-
based anodes in lithium-ion batteries.
In our previous studies,6-7 a thermal vapor decomposition ͑TVD͒
technique has been applied to coat carbon onto the surface of natural
graphite particles so as to protect the graphite ‘‘core’’ from the con-
tact with PC-based electrolytes. It was found that the TVD carbon
coating can suppress the decomposition of most electrolytes to a
very small extent at low voltage. So it is assumed that the TVD
technique is also applicable to the case of Si.
In this study, carbon-coated Si has been prepared by TVD. The
excellent electrochemical performance of carbon-coated Si as an
anode material for lithium-ion batteries has been demonstrated.
Experimental
Prior to the TVD carbon-coating process, Si powder ͑Wako, av-
erage particle size 8 ␮m͒ was ground in a disk mill for 1 h and then
sieved. Si powder with particle size smaller than 53 ␮m was used
for TVD carbon-coating treatment. 10 g of the obtained Si powder
was put into a quartz tube furnace purged with a N2 stream carried
with benzene vapor ͑flow rate, 1 L/min͒. The temperature of the
furnace was kept at 1000°C. At such a high temperature, the ben-
Table I. Some physical properties of the carbon-coated Si
sample.
Weight percents of carbon coating
Average particle size
Tap density
20%
18 ␮m
0.9 g/cm3
2.8 m2/g
95 m⍀ cm
BETa specific surface area
Volume resistanceb
a Brunauer-Emmett-Teller.
b Volume resistance of the Si powder before TVD treatment is 1500 ⍀
cm.
* Electrochemical Society Active Member.
z E-mail: yoshio@ccs.ce.saga-u.ac.jp
Downloaded on 2013-02-24 to IP 137.99.31.134 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

Journal of The Electrochemical Society, 149 ͑12͒ A1598-A1603 ͑2002͒
A1599
Figure 1. X-ray diffraction ͑XRD͒ pattern of carbon-coated Si powder.
rested for 10 min; (iii) the cell was discharged at a constant current
Figure 3. Typical charge/discharge curves of the carbon-coated Si/Li metal
cells using the constant capacity charge mode in the electrolyte of 1 M
LiPF6-EC:DC ͑1:2 by volume͒.
density of 0.5 mA/cm2 to 1.5 V; (i ) the cell was rested for 10 min;
v
then the cycles were repeated. For carbon-coated Si/LiMn2O4 ,
carbon-coated Si/LiCoO2 , and carbon-coated Si/LiNiO2 cells, the
cells were generally charged to 800 mAh/g vs. Si-based electrodes
and then discharged to 2.5 V at the same current density ͑there was
10 min of rest after each cycle charge or discharge͒. By contrast, in
the second set, namely, ‘‘constant voltage charge’’ mode, the proce-
dures for carbon-coated Si/Li metal cell was as follows: (i) the cell
was discharged/charged by the ‘‘constant capacity charge’’ mode in
the first cycle; (ii) from the second cycle on, the cell was charged
from 1.5 V to 100 mV at the constant current density of 0.5 mA/cm2
and held at this voltage for 130 min; (iii) the cell was rested for 10
Figure 4. Typical charge/discharge curves of the carbon-coated
Si/LiMn2O4 , carbon-coated Si/LiCoO2 , and carbon-coated Si/LiNiO2 cells
using the constant capacity charge mode in the electrolytes of 1 M
LiPF6-EC:DMC ͑1:2 by volume͒.
Figure 2. Typical CVs of carbon-coated Si in the electrolytes of 1 M
LiPF6-EC:DMC ͑1:2 by volume͒ at the scan rates of ͑a͒ 1 mV/s; and ͑b͒ 0.1
mV/s.
Downloaded on 2013-02-24 to IP 137.99.31.134 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

A1600
Journal of The Electrochemical Society, 149 ͑12͒ A1598-A1603 ͑2002͒
Figure 5. Cycle life behavior of carbon-coated Si using the constant voltage
charge mode in the electrolyte of 1 M LiPF6-EC:DMC ͑1:2 by volume͒.
Figure 6. Initial charge/discharge curves of the carbon-coated Si/LiCoO2
cell using constant capacity charge ͑520 Ah/g͒ mode in the electrolytes of 1
M LiPF6-EC:DMC ͑1:2 by volume͒ and 1 M LiPF6-PC.
min; (i ) the cell was discharged with the constant current density
v
of 0.5 mA/cm2 to 1.5 V; then the cycles were repeated from steps ii
to iv.
In the CV experiments, three-electrode cells were used. In these
cells, lithium metal electrodes were used as both the counter and
reference electrodes, and the Si-based electrodes were used as the
working electrodes. The cells were cycled between OCV ͑ca. 3 V vs.
Li/Liϩ) and 0 V vs. Li/Liϩ at different scan rates.
In order to evaluate the thermal safety issue of carbon-coated Si
as an anode material for lithium-ion batteries, we performed differ-
ential scanning calorimetry ͑DSC͒ studies on lithiated carbon-coated
Si. DSC of lithiated MCMB-6-28 ͑Osaka Gas͒ is also investigated
as a reference. The samples were washed with the solvent of DMC
and dried over vacuum at room temperature prior to placement into
DSC pans.
To gain more insights into our studies on carbon-coated Si, the
Si-based electrodes were charged and discharged to different extents
of lithiation and examined by 7Li nuclear magnetic resonance
͑NMR͒ spectroscopy. The details are similar to those described in
our previous paper.6
every Si-based particle and few lithium ions can penetrate into the
Si core since the diffusion rates in both the Si and carbon phases are
too slow to catch up with the potential sweep. As the scan rate
decreases to certain degree, like 0.1 mV/s, more and more lithium
ions can follow up with the potential steps and enter the Si core.
Thus the peaks due to Li-Si alloying/dealloying predominate in the
CV. As shown in the CV at 0.1 mV/s, the humps range from 0.7 to
1.3 V vs. Li/Liϩ. In the CV at 1 mV/s, the humps can hardly be seen
in both cathodic and anodic directions. The humps are fingerprints
of a carbon phase. By contrast, there are roughly three anodic peaks
that can be clearly observed at about 0.20, 0.36, and 0.53 V vs.
Li/Liϩ, respectively. A shoulder at about 0.3 V vs. Li/Liϩ can also
be watched at the negative-potential side of the anodic peak at
0.36 V vs. Li/Liϩ. These delicate peaks originate from the phase
transitions between Li4.4Si, Li3.25Si, Li2.33Si, Li1.71Si, and Si,
respectively.9 On the other hand, the corresponding delicate anodic
peaks can barely be observed in the cathodic scans of the CV at 0.1
mV/s. Instead, the cathodic current begins to rise up slopingly from
ca. 0.3 V vs. Li/Liϩ until the lowest potential of 0 V vs. Li/Liϩ,
which may come from the big polarization due to low diffusion of
lithium ions in Si phase. Comparison of the CVs at 1 and 0.1 mV/s
also indicates the difference in the cyclabilities of the carbon-coating
and Si core in the potential range from OCV to 0 V vs. Li/Liϩ. In
the CV at 1 mV/s, there is little decrease of the currents with cy-
cling, which implies the good cyclability of carbon coating phase.
On the contrary, noticeable decreases can be seen in the currents
during cycling in the CV at 0.1 mV/s. This means that the capacity
is fading in Si phases along with cycling.
Figures 3 shows the typical charge/discharge curves of the
carbon-coated Si/Li metal cells by using the ‘‘constant capacity
charge’’ mode. In the first cycle, the charge capacity ͑intercalation of
lithium into Si-based electrode͒ was 800 mAh/g while the discharge
capacity ͑deintercalation from Si-based electrode͒ was 737 mAh/g.
Thus the coulombic efficiency in the first cycle was 92.1%. From
the second until the 50th cycle, the values of discharge and charge
capacity are very near for every cycle, which implied very high
coulombic efficiency ͑near 100%͒. It is well known that Li alloys
with Si with the compositions of Li1.71Si, Li2.33Si, Li3.25Si, and
Li4.4Si at elevated temperature.9 In our studies, it is anticipated that
if the charge capacity is restricted less than the corresponding ca-
pacity of Li1.71Si ͑1640 mAh/g-Si͒, the big expansion of Si lattice
due to phase transition can be considerably suppressed and thus the
cyclability of Si-based electrode is improved.10
Results and Discussion
Figure 1 shows the X-ray diffraction ͑XRD͒ pattern of carbon-
coated Si powder. The peaks due to carbon and Si diffractions can
be clearly observed, whereas no peak corresponding to SiC can be
seen. In the previous study of Wilson et al., SiC is formed in car-
bons containing nanodispersed Si atoms if the content of silicon
exceeds 11 atom %.8 SiC is electrochemically inactive toward
lithium, which will decrease the overall capacity of the composite
electrode material. Although both of the studies included gaseous
sources for preparing the Si-C composite anode material, we only
used organic vapor for the carbon-coating source but adopted Si
powder as a starting material, whereas Wilson et al. utilized gases as
all the starting materials. This may be one of the reasons in the
difference of the products.
To get a rough estimation of the electrochemical behavior for the
carbon-coated Si, cyclic voltammetric experiments were conducted
at first. Figure 2 shows the typical cyclic voltammograms ͑CV͒ of
carbon-coated Si at the scan rates of 1 and 0.1 mV/s, respectively.
The three-electrode CV cells were cycled at high scan rate ͑1 mV/s͒
prior to slow scan rate ͑0.1 mV/s͒. In the first cycle of CV, the
irreversible peak of electrolyte decomposition ranging from 0.5 to
1.5 V vs. Li/Liϩ can hardly be seen ͑all the CV curves in this
potential range overlap with each other throughout the fifth cycles͒,
which implies that the electrolyte decomposition on the surface of
carbon coating can be suppressed to a very small extent, and thus
ensure high reversibility of this material. Moreover, in the CV at
high scan rate, say, 1 mV/s, the features due to lithium intercalation/
deintercalation into or from the carbon coating are mainly demon-
strated, as discussed in Ref. 7. It is expected that at high scan rates,
most lithium ions only diffuse in the outer carbon-coating layer of
Figure 4 shows the typical charge/discharge curves of the
carbon-coated Si/LiMn2O4 , carbon-coated Si/LiCoO2 , and carbon-
coated Si/LiNiO2 cells using the constant capacity charge mode. The
Downloaded on 2013-02-24 to IP 137.99.31.134 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

Journal of The Electrochemical Society, 149 ͑12͒ A1598-A1603 ͑2002͒
A1601
Figure 7. 7Li NMR of the Si-based electrodes with different charge and
discharge capacities, ͑a͒ charge capacity of 500 mAh/g in the first and fifth
cycle using constant capacity charge ͑500 mAh/g͒ mode; ͑b͒ charge capaci-
ties of 500, 1000, and 2000 mAh/g in the first cycle using ‘‘constant capac-
ity charge’’ mode; ͑c͒ discharge capacities of 500, 1000, and 2000 mAh/g.
Downloaded on 2013-02-24 to IP 137.99.31.134 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

A1602
Journal of The Electrochemical Society, 149 ͑12͒ A1598-A1603 ͑2002͒
weight ratios between the cathode materials (LiMn2O4 , LiCoO2 ,
and LiNiO2) and carbon-coated Si anode are indicated in the figure.
In these cells, the high coulombic efficiency as in the carbon-coated
Si/Li metal cell could not be obtained and the discharge capacities
͑deintercalation from Si-based anodes͒ were a little lower than 800
mAh/g. This may be due to some limitations in cathode materials.
Moreover, the average working voltages of these cells were ca. 3.5
V ͑carbon-coated Si/LiMn2O4), 3.4 V ͑carbon-coated Si/LiCoO2),
and 3.25 V ͑carbon-coated Si/LiNiO2), respectively. These values
are ca. 0.3 V lower than the working voltages of the corresponding
cells using graphite as anode material.
As demonstrated above, high reversible capacity of 800 mAh/g
can be obtained by the constant capacity charge mode for carbon-
coated Si. Figure 5 shows the cycle life behavior of carbon-coated Si
using the constant voltage charge mode. After being charged to 800
mAh/g and discharged to 1.5 V in the first cycle, the cell was
charged to 100 mV and held at this voltage for 130 min from the
second cycle on. As a result, both the charge and discharge capaci-
ties in the second cycle are low, and from the third to the 10th cycle,
both the charge and discharge capacities grow up until the saturation
value of 980 mAh/g. From the 10th cycle, the capacities fade slowly
with the cycles. Therefore it seems that charge ͑intercalation of
lithium into Si-based electrode͒ mode plays a very important role in
the cyclability of the carbon-coated Si electrode.
The above satisfactory electrochemical performance of carbon-
coated Si was obtained in the EC-based electrolyte of 1 M
LiPF6-EC:DMC ͑1:2 by volume͒. To make carbon-coated Si a ver-
satile anode material for lithium-ion batteries, the compatibility of
carbon-coated Si with PC-based electrolyte was also tested. Figure 6
shows the initial charge/discharge curves of the carbon-coated
Si/LiCoO2 cell in the electrolytes of 1 M LiPF6-EC:DMC ͑1:2 by
volume͒ and 1 M LiPF6-propylene carbonate ͑PC͒. It can be seen
that there is almost no difference in the charge curves for both the
EC and PC-based electrolytes, other than one small shoulder at 3.2
V in the charge curve for 1 M LiPF6-PC electrolyte, which is due to
the mild decomposition of the electrolyte on the surface of carbon-
coated Si. This indicates that a Si-based electrode also is compatible
with PC-based electrolytes.
Figure 8. DSC thermal profiles of carbon-coated Si lithiated to 800 mAh/g
and MCMB 6-28 lithiated to 330 mAh/g.
0 ppm can be observed. The intensities of this signal are very high.
So it seems more likely the peak at 0 ppm originates from the Liϩ
irreversibly bonded with Si-based material. With increasing lithia-
tion extent ͑charge capacities of 500, 1000, and 2000 mAh/g͒, as
shown in Fig. 7b, the intensities of the peaks at 7.3 and 20.2 ppm
increase whereas that at 0 ppm decreases relatively, and becomes a
shoulder of the former peaks. In the 7Li NMR spectra of carbon-
coated Si electrodes fully delithiated from the charged capacities of
500, 1000, and 2000 mAh/g, respectively, as shown in Fig. 7c, the
signal at 0 ppm is always very large, and the signal at 7.3 ppm
becomes a very small shoulder of the peak at 0 ppm in the cases of
charge capacities of 1000 and 2000 mAh/g. The signal at 20.2 ppm
cannot be seen in all the spectra of fully delithiated Si-based elec-
trodes. Therefore, the signal at 0 ppm can be mainly attributed to
Liϩ irreversibly bonded with carbon-coated Si, whereas the signal at
20.2 ppm originates from lithium reversibly alloyed with Si.
To evaluate the thermal stability of carbon-coated Si electrodes,
DSC experiments were carried out on lithiated carbon-coated Si and
mesocarbon microbeads ͑MCMB͒ 6-28. Figure 8 shows the DSC
thermal profiles of carbon-coated Si lithiated to 800 mAh/g and
MCMB 6-28 lithiated to 330 mAh/g. The exothermic peaks lower
than 150°C can be observed for both lithiated Si-based material and
MCMB, which originate from the breakdown of the SEI layer.11,12
These peaks demonstrate close intensities. This implies similar exo-
thermic heat generations. Moreover, the onset temperature of the
exothermic peak corresponding to lithiated carbon-coated Si is a
little higher than that of the lithiated MCMB. In addition, two exo-
thermic peaks appear at temperatures higher than 250°C for each
DSC thermal profile. These peaks may be due to the reaction of
intercalated lithium ͑into Si or C͒ with PVDF binder.12 The exother-
mic heat generation corresponding to the peaks of lithiated carbon-
coated Si higher than 250°C amount to almost three times of that of
lithiated MCMB. This seems in accord with their relative capacities
to accommodate lithium. Thus, carbon-coated Si is a little thermally
safer than the conventional anode material, graphitic carbon, at tem-
peratures lower than 150°C.
To shed more light on the storage states of lithium in Si-based
electrodes, 7Li NMR was applied on the Si-based electrodes with
different charge and discharge capacities as shown in Fig. 7. In the
7Li NMR spectra ͑as shown in Fig. 7a͒ of the carbon-coated Si
electrode being charged ͑intercalation of lithium͒ with 500 mAh/g in
the first cycle, there are mainly three peaks at the chemical shifts of
Ϫ2.7, 4.7, and 17.7 ppm, respectively. On the other hand, in the 7Li
NMR spectra of carbon-coated Si electrode charged with 500 mAh/g
in the fifth cycle ͑after constant capacity charge of 500 mAh/g for 4
cycles͒, two peaks can be observed at the chemical shifts of 7.3 and
20.2 ppm, respectively. If we correct the chemical shift of the peak
7
at Ϫ2.7 to 0 ppm in the former Li NMR spectra, the peaks at 4.7
and 17.7 ppm can be modified to 7.4 and 20.4 ppm, repectively. The
7Li NMR spectra of carbon-coated Si lithiated to 500 mAh/g at
different cycles ͑first and fifth͒ demonstrate almost the same signals
at ca. 7.3 and 20.2 ppm, except the peak at 0 ppm in the spectra at
the first cycle. Moreover, the intensity ratios of the peaks at 7.3 and
20.2 ppm for both spectra are near. From the chemical shifts of the
7
signals in the Li NMR spectra of lithiated Si-based electrodes, the
Conclusion
highly ionic states of lithium stored in the carbon-coated Si can be
clarified. As for the peak at 0 ppm in the first cycle, it may originate
from two sources: ͑1͒ the signal of the residue electrolyte on the
surface of the Si-based electrode; ͑2͒ some response from lithium
inserted into Si based electrodes. However, the signal from the resi-
due electrolyte must be much weaker as compared with the response
from lithium inserted into the electrode since the electrode material
was tested by 7Li NMR after being rinsed with DMC solvent and
So far we have showed the excellent electrochemical perfor-
mance of carbon-coated Si as anode materials for lithium-ion bat-
teries in terms of high reversible capacity over 800 mAh/g, high
coulombic efficiency, good cyclability, satisfactory compatibility
with both the EC and PC-based electrolytes, and better thermal sta-
bility than that of graphite, etc. Carbon-coating in the outer layer
plays a very important role in the improvement of the electrochemi-
cal behavior. It could not only considerably suppress the big decom-
position of electrolytes on the surface of Si-based electrodes, but
also provide integral and continuous electric contact networks
7
dried. In the Li NMR spectra of the carbon-coated Si electrodes in
fully discharged ͑deintercalation of lithium from 500 mAh/g lithi-
ated carbon-coated Si states, as shown in Fig. 7c, only the signal at
Downloaded on 2013-02-24 to IP 137.99.31.134 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

Journal of The Electrochemical Society, 149 ͑12͒ A1598-A1603 ͑2002͒
A1603
3. H. Li, X. Huang, L. Chen, Z. Wu, and Y. Liang, Electrochem. Solid-State Lett., 2,
547 ͑1999͒.
4. I. Kim, P. N. Kumta, and G. E. Blomgren, Electrochem. Solid-State Lett., 3, 493
͑2000͒.
5. S. Hwang, H. Lee, S. Jang, S. Lee, S. Lee, H. Baik, and J. Lee, Electrochem.
Solid-State Lett., 4, A97 ͑2001͒.
6. M. Yoshio, H. Wang, K. Fukuda, Y. Hara, and Y. Adachi, J. Electrochem. Soc., 147,
1245 ͑2000͒.
7. H. Wang and M. Yoshio, J. Power Sources, 93, 123 ͑2001͒.
8. A. M. Wilson and J. R. Dahn, J. Electrochem. Soc., 142, 326 ͑1995͒.
9. B. A. Boukamp, G. C. Lesh, and R. A. Huggins, J. Electrochem. Soc., 128, 725
͑1981͒.
around Si particles even they are a little expanded after lithium
insertion. 7Li NMR spectra of lithiated/delithiated Si based elec-
trodes demonstrates three kinds of storage sites for lithium in
carbon-coated Si, corresponding to the signals at 0, 7.3, and 20.2
ppm, respectively. The signals at 0 and 20.2 ppm can be ascribed to
lithium irreversibly bound with and reversibly alloyed with Si, re-
spectively. Lithium-ion batteries using carbon-coated Si as anode
materials have average working voltages about 0.3 V lower than
those using graphite.
Saga University assisted in meeting the publication costs of this article.
10. T. Umeno, K. Fukuda, H. Wang, N. Dimov, T. Iwao, and M. Yoshio, Chem. Lett.,
2001, 1186.
References
11. M. N. Richard and J. R. Dahn, J. Electrochem. Soc., 146, 2068 ͑1999͒.
12. A. Du Pasquier, F. Disma, T. Bowmer, A. S. Gozdaz, G. Amatucci, and J.-M.
Tarascon, J. Electrochem. Soc., 145, 472 ͑1998͒.
1. J. O. Besenhard, J. Yang, and M. Winter, J. Power Sources, 68, 87 ͑1997͒.
2. C. S. Wang, G. T. Wu, X. B. Zhang, Z. F. Qi, and W. Z. Li, J. Electrochem. Soc.,
145, 2751 ͑1998͒.
Downloaded on 2013-02-24 to IP 137.99.31.134 address. Redistribution subject to ECS license or copyright; see www.esltbd.org