High-rate lithium ion batteries with flat plateau based on self-nanoporous structure of tin electrode

Article abstract of DOI:10.1149/1.2405727

Tin film as a negative electrode in Li-ion secondary batteries with a flat plateau in the potential vs charge relationship even at a high rate (high current density) were fabricated. The self-nanoporous structure, which means that electrode becomes nanoporous after charge/discharge cycling, due to the lithium insertion/extraction during high-rate charge/discharge, high lithium diffusion coefficients, and metallic conductivity based on tin and tin-lithium alloy, are suitable for the high-rate property. The flat plateau of charge/discharge behavior indicates that this lithium storage device is not a super-capacitor but a secondary battery. This is the first report of a high-rate lithium secondary battery with a flat plateau.

Full text of DOI:10.1149/1.2405727

A146
Journal of The Electrochemical Society, 154 ͑2͒ A146-A149 ͑2007͒
0
013-4651/2006/154͑2͒/A146/4/$20.00 © The Electrochemical Society
High-Rate Lithium Ion Batteries with Flat Plateau Based
on Self-Nanoporous Structure of Tin Electrode
a
a
a
b
Eiji Hosono, Hirofumi Matsuda, Itaru Honma, Masaki Ichihara, and
a,z
Haoshen Zhou
a
National Institute of Advanced Industrial Science and Technology, Energy Technology Research Institute,
Tsukuba 305-8568, Japan
b
Material Design and Characterization Laboratory, Institute for Solid State Physics, University of Tokyo,
Kashiwa, Chiba 277-8581, Japan
Tin film as a negative electrode in Li-ion secondary batteries with a flat plateau in the potential vs charge relationship even at a
high rate ͑high current density͒ were fabricated. The self-nanoporous structure, which means that electrode becomes nanoporous
after charge/discharge cycling, due to the lithium insertion/extraction during high-rate charge/discharge, high lithium diffusion
coefficients, and metallic conductivity based on tin and tin–lithium alloy, are suitable for the high-rate property. The flat plateau
of charge/discharge behavior indicates that this lithium storage device is not a super-capacitor but a secondary battery. This is the
first report of a high-rate lithium secondary battery with a flat plateau.
©
2006 The Electrochemical Society. ͓DOI: 10.1149/1.2405727͔ All rights reserved.
Manuscript submitted May 30, 2006; revised manuscript received September 26, 2006.
Available electronically December 27, 2006.
Lithium-ion storage devices have been widely studied for pos-
sible application in electric devices, especially for mobile or por-
table electric devices and electric vehicles, which require high en-
ergy density and high power density. The synthesis of electrode
materials for lithium-ion storage devices with high specific capacity
at high charge/discharge current rate has been expected because
lithium-ion storage devices with high specific capacities ͑ϭhigh en-
ergy density͒ and high current densities ͑ϭhigh power density͒ are
necessary for industrial use. Many papers have reported materials
with only high specific capacities, using nanocrystalline transition
vices cannot be called high-rate lithium-ion secondary batteries but
should rather be called high-rate lithium storage devices or superca-
pacitors. Such high-rate lithium storage devices or super-capacitors
cannot satisfy industrial needs, so there is a need to develop high-
rate lithium-ion secondary batteries with potential flat plateaus in the
charge/discharge process.
Here, the metallic tin on SUS-304 mesh is used as a negative
electrode for the high-rate lithium-ion secondary batteries with po-
tential flat plateaus, which are fabricated by electrodeposition, be-
cause tin has a high theoretical capacity of 994 mAh/g and high
electronic conductivity based on the metallic conductivity and high
1
,2
3,4
5-10
metal oxides or the alloy reaction such as Si and Sn.
Re-
−
7
−8
−2 −1 16
cently, reports of electrodes with both high specific capacities and
diffusion coefficient of Li–Sn alloy of 10 –10 cm s , which
is over 1000 times the diffusion coefficient of general metal oxide
11-15
high current densities have increased.
However, high energy and
electrodes of 10− –10 cm s .
10
−17
−2 −1 13,17
So, there are not only four
high power lithium-ion secondary batteries with a flat plateau of the
charge/discharge curves have not been reported.
problems to be solved for high-rate lithium-ion batteries; a fifth
problem is how to increase the lithium-ion diffusion coefficient in
active materials to decrease the diffusion polarization in the high-
rate charge/discharge process. This paper is the first to report the
synthesis of nanoporous metal tin as an anodic electrode for high-
rate lithium-ion secondary batteries with a flat plateau and good
cycle performance. The properties of the high-rate secondary batter-
ies of metallic tin include good metallic conductivity, high lithium-
ion diffusion coefficient, and self-porous reaction process due to the
lithium insertion/extraction by high-rate charge/discharge.
Generally, there are the following four problems which arise due
to the nature of lithium-ion batteries or lithium storage devices,
which need to be solved in order to develop high rate lithium storage
devices: ͑i͒ increasing the electronic conductivity of the electrode
materials by using highly conductive materials, ͑ii͒ reducing the
required diffusion length in the active materials by decreasing the
particle size, ͑iii͒ reducing the effective specific current density by
high surface area, and ͑iv͒ realizing high cycle performance by
nanoparticles and porous structure to ease the tension caused by
volume expansion upon insertion of the lithium.
Experimental
In our recent work, lithium storage devices with both high power
and energy densities by fabricating a nanocrystalline and mesopo-
rous metal oxide film on a nickel mesh, knitted of nickel
A precursor solution was prepared by dissolving SnCl ·2H O
2
2
͑
99.9% purity, Wako Pure Chemicals Co., Ltd., Japan͒ in H O with
11,12
2
micrometer-wires via the self-template method, are reported.
HCl ͑99.9% purity, Wako Pure Chemicals Co., Ltd., Japan͒, which
was added dropwise to the solutions to adjust pH values ͑about 0.9͒.
The novel structure partly mitigates the above four problems and
enables a high specific capacity to be obtained at a high charge/
discharge current rate. However, the thickness of the metal oxide
layer on the Ni micrometer wire should be less than several hundred
nanometers. The thickness of the metal oxide layer cannot be in-
creased in order to obtain high electronic conductivity because the
resistivity of metal oxide ͑NiO11 or Fe O ͒ is high. Hence, the
−3
The concentration of the tin ions was fixed at 0.0075 mol dm .
The cell for deposition was a conventional three-electrode cell in
which a Pt͑200 nm͒/Ti/glass plate was used as counter electrode
with separation from the working electrode of SUS-304 ͑100 mesh͒
for tin deposition. A saturated calomel electrode ͑SCE͒ was used as
the reference electrode. A constant potential of −2.0 V was applied
to the SUS-304 for 5 min in the precursor solution. The deposited
tin film was rinsed by deionized water and dried at room tempera-
ture in the vacuum condition.
12
3
2
weight ratio of active materials is very small. Moreover, the greatest
weakness of high-rate lithium storage devices reported so far is the
profile of the charge/discharge potential curve, which is a capacitor-
8
,11-15
like curve at the high-rate charge/discharge condition.
In the
Electrochemical measurements were carried out using the three
electrodes set up in a twin beaker cell connected with a microcapil-
lary acting as the separator. The fabricated tin film on the mesh was
used as the working electrode. The reference and counter electrode
were prepared by spreading and pressing of lithium metals on the
capacitor-like curves, the charge/discharge curves show no potential
flat plateaus and a linear decrease/increase of the potential in accor-
dance with the lithium insertion/extraction. Hence, the reported de-
SUS-304 mesh ͑100 mesh͒. A 1 mol dm− LiClO in ethylene car-
3
4
z
E-mail: hs.zhou@aist.go.jp
bonate ͑EC͒/diethyl carbonate ͑DEC͒ was used as electrolyte. Cell
Downloaded on 2015-02-11 to IP 128.122.253.228 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 154 ͑2͒ A146-A149 ͑2007͒
A147
Figure 2. ͑Color online͒ ͑a͒ Constant current discharge/charge curves of the
2
nd cycle at 0.1 and 1.0 A/g and ͑b and c͒ constant current discharge/charge
−
3
curves at 10 and 20 A/g, respectively. Those were recorded in 1 mol dm
LiClO in EC/DEC electrolyte.
4
The properties of the lithium-ion secondary batteries using tin
anodes were tested via constant current charge/discharge measure-
ment. Figure 2a shows the second charge/discharge cycling curves
performed at 0.1 and 1 A/g over the voltage range of 0.01–0.8 V vs
+
a Li/Li reference electrode. In the curves at 0.1 A/g, the discharge
Figure 1. SEM images of the tin films on SUS-304 fabricated by the elec-
trodepostion at −2.0 V ͑vs SCE͒ for 5 min.
assembly was carried out in a glove box under an argon atmosphere.
The charge/discharge performance of the materials was investigated
in such a three-electrode cell using lithium metal as counter and
reference electrodes. The weight in specific capacity ͑mAh/g͒ and
current rate ͑A/g͒ was calculated by only active materials. The
charge/discharge performance was measured at various current den-
sities of 0.1, 1, 10, and 20 A/g. The potential window at the low-rate
condition ͑0.1 and 1 A/g͒ and at the high-rate condition ͑10 and
20 A/g͒ was 0.01–0.8 and 0.01–1.5 V, respectively.
Results and Discussion
Figure 1 shows the electrodeposited tin film on the SUS-304
mesh. The tin dendrite grown on the SUS-304 can be seen in the low
magnification images in Fig. 1a and b. The length of tin dendrite is
around several tens of micrometers. Figure 1c shows tin particles
with sizes ranging from several hundred nanometers to several mi-
crometers; thus, each dendrite is not a nanostructure but a micro-
structure. The morphology of each dendrite is similar to that of
Figure 3. ͑Color online͒ ͑a͒ XRD patterns of the tin film before the charge/
discharge measurement and ͑b͒ after the 21st cycle charge/discharge mea-
surement at 10 A/g.
18
electrodeposited dendrite.
Downloaded on 2015-02-11 to IP 128.122.253.228 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A148
Journal of The Electrochemical Society, 154 ͑2͒ A146-A149 ͑2007͒
Figure 4. ͑a–c͒ SEM images and ͑d͒ TEM image of the porous tin film after
the 21st cycle charge/discharge measurement at 10 A/g.
Figure 5. ͑Color online͒ ͑a͒ Constant current discharge/charge curves at
+
1
.0 A/g over the voltage range of 0.3–0.7 V vs a Li/Li reference electrode
and ͑b͒ constant current discharge/charge curves at 20 A/g over the voltage
+
range of 0.9–0.14 V vs a Li/Li reference electrode.
curve indicates three plateaus from 0.7 to 0.4 V. After the third
plateau, the curve gradually decreases to 0.01 V. The discharge
curve is similar to that reported previously.
7
The first plateau at 0.7 V is the reaction
which is near to 300 mAh/g at the low-rate condition of 0.1 A/g.
The theoretical capacity of LiSn ꢀ Li7/3Sn is 301 mAh/g. Hence,
the flat plateau capacity of around 250 mAh/g even at high rates
such as 10 and 20 A/g is near to the theoretical capacity. This flat
plateau with good cycle performance at the high-rate condition is
suitable for high-rate secondary batteries. In many reports, tin-based
Sn → Li0.4Sn
The second plateau at around 0.55 V is the reaction
Li0.4Sn → LiSn
͓1͔
͓2͔
͓3͔
͓4͔
The third plateau at around 0.4 V is the reaction
9,19
6
intermetallic compounds such as Cu–Sn
and Sb–Sn were used
for the anode electrode because of the poor cyclability of pure tin
LiSn → Li7/3Sn
7
due to the large volume change. However, our results show that the
The gradual decreasing curve is the reaction
Li7/3Sn → Li2.5Sn → Li2.6Sn → Li3.5Sn → Li4.4Sn
pure tin anode electrode had good cycle performance at the high-rate
condition.
First, we discuss the flat plateau at the high-rate condition. It is
considered that solving problems 1, 2, 3, and 5, which is the use of
materials having high lithium-ion diffusion coefficients, results in
good charge/discharge curves. The solution of problem 1, which is
electron conductivity, is confirmed from the X-ray diffraction
In this case, the capacity of the second cycle at 0.1 A/g is about
Li3.5Sn.
It is considered that the typical lithium insertion reaction is
caused by this electrodeposited tin film on SUS-304 because the
discharge curve clearly shows almost all of Reactions 1-4. In the
charge/discharge curves at the relatively high rate of 1.0 A/g, the
plateau at 0.7 V in discharge, which is observed in the cycling
curves at 0.1 A/g, is converted into the drastic decreasing curve
with lithium insertion, like a capacitor. The large capacity of around
80 mAh/g includes the plateau length with 250–300 mAh/g at
.4 V in discharge, corresponding to the lithium insertion from
LiSn–Li7/3Sn, which is similar to that in the cycling curves at
͑
XRD͒ patterns ͑Fig. 3͒ of the electrodeposited tin film before
charge/discharge measurement and after the 21st cycle at 10 A/g.
Judging from the XRD patterns, which indicate the metallic tin pat-
tern, the metallic tin is maintained after the lithium insertion/
extraction reaction. The active material is always metallic during the
charge/discharge from start to end because the mechanism of lithium
6
0
7
storage for the tin electrode is the tin-lithium alloy reaction. The
maintenance of the metallic conductivity is suitable for high-rate
lithium-ion cells. The solution of problems 2 and 3 based on nano-
structure is confirmed by the scanning electron microscopy ͑SEM͒
and transmission electron microscopy ͑TEM͒ images of the tin films
before and after the 21st cycle at 10 A/g in Fig. 1 and 4, respec-
tively. From the low magnification image in Fig. 4a and b, the den-
drite is not removed from the SUS-304 mesh. The thickness of the
tin film is around 5–10 ␮m ͑Fig. 4b͒, which is thicker than that
before charge/discharge as shown in Fig. 1b. As compared with the
SEM images between Fig. 1c and 4c, the dendrite tin film is con-
verted into the nanoporous structure. The self-nanoporous reaction
based on the volume-change reaction of Li–Sn alloy as shown in
Fig. 4c results in increased film thickness. The thickness of
0
.1 A/g. Here, the cycle performance such as the 20th cycle is poor
due to the large volume change of the lithium insertion from Sn to
7
Li4.4Sn.
Figures 2b and c show the charge/discharge cycling curves per-
formed at the high rate of 10 and 20 A/g over the voltage range of
+
0
.01–1.5 V vs a Li/Li reference electrode, respectively. Taking into
account the overpotential due to the increase of current density re-
sulting in the change of voltage range, the plateaus at 0.4 V in the
discharge curve at the low rate ͑0.1 A/g͒ decrease to 0.15 V in both
the 10 and 20 A/g curves. The first and second plateaus at 0.7 and
0
.55 V in the cycling curves at 0.1 A/g in Fig. 2a are converted into
the capacitor-like curve. The curves from the second cycle to 20th
cycle at 10 and 20 A/g indicate good cycle performance. Each dis-
charge capacity shows a large capacity of around 350–400 mAh/g,
which is similar to that of graphitic carbon, in spite of the very-high-
rate condition. Moreover, the most important characteristic is the
stable plateau, which is derived from the reaction of LiSn
11,12
5–10 ␮m is much larger than the thickness of metal oxides
on
Ni wire for high-rate lithium-ion storage devices. This nanoporous
structure is constructed by the nanoparticles as shown in TEM im-
ages in Fig. 4d. This nanoporous structure, which is created by self-
reaction in the high-rate insertion/extraction based on the volume
change reaction of Li–Sn alloy, decreases the required diffusion
length in the active materials and the effective specific current den-
→
Li7/3Sn at around 0.15 V at the high-rate condition of 10 and
0 A/g. The plateau length with around 250 mAh/g still remains,
2
Downloaded on 2015-02-11 to IP 128.122.253.228 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 154 ͑2͒ A146-A149 ͑2007͒
A149
Table I. Comparison of the capacity in the limited potential range as the lithium-ion secondary batteries. We use the ratio of capacity in per
potential unit, which is defined as CЈ/⌬V, to estimate the lithium storage device and supercapacitor, here CЈ is the capacity and ⌬V is the limited
potential range.
Current in
the cell
͑mA͒
CЈ in
limited region
͑mAh/g
Rate
⌬V
͑V͒
CЈ/⌬V
−1
͑mAh/g V−1
−1
Material ͑Ref.͒
͑Ag−1
͒
͒
͒
Classify
This work
10.0
15.1
19.0
10.0
332
350
225
0.15–0.01 V
0.20–0.01 V
1.8–1.0 V
2370
1842
280
Secondary batteries
20.0
TiO –P O –SnO
10.0
Supercapacitor
2
2
5
2
1
5
CGMN
1
4
c-SWNT/TiO ͑TTB-acac͒
6.72
23.7
10
—
1.9
8
60
560
540
700
1.7–1.4 V
0.9–0.34 V
1.3–0.5 V
1.0–0.01 V
200
1000
675
Supercapacitor
Supercapacitor
Supercapacitor
Supercapacitor
2
8
SnO nanofibers
2
11
Ni/NiO
1
2
Ni/Fe O
13
11
707
2
3
8
sity. The solution of problem 5 is based on the high lithium-ion
diffusion coefficient of Li–Sn alloy. The lithium-ion diffusion coef-
ever, the absolute current of 1.9 mA in the cell for 23.7 A/g is
much smaller than those of other works due to the very small
amount of active materials. From Table I, only our device can be
called a lithium-ion secondary battery.
−
7
2
−1
16
ficient of 1.8–5.9 ϫ 10 cm s of Li7/3Sn is extremely large.
8,11-15
Previous reports
on high-rate lithium storage devices did not
consider the use of materials with high lithium-ion diffusion coeffi-
cients. Moreover, the electronic conductivity of the active materials
in the reports was lower than our metallic conductivity because the
other materials were metal oxides. In this work, the electronic con-
ductivity and lithium-ion diffusion coefficient are vastly improved,
thus attaining the flat plateau at the high-rate condition.
Conclusions
Electrodeposited tin dendrite indicates a large capacity of
20 mAh/g ͑10th cycle͒ based on the large flat plateau at high cur-
4
rent densities of 10 and 20 A/g. The metallic electron conductivity,
high lithium-ion diffusion coefficients, and self-nanoporous struc-
ture of tin during high-rate charge/discharge due to tin–lithium alloy
are suitable for the flat plateau at the high-rate condition. Moreover,
the small volume change based on the Sn–Li7/3Sn reaction results in
good cycle performance. The lithium-ion batteries based on this me-
tallic tin as the negative electrode are not super-capacitors but real
high-rate lithium secondary batteries with a flat plateau and good
cycle performance.
Second, the solution of problem 4, which is cycle performance,
at the high-rate condition is discussed. In a previous study using a
7
tin electrode, the cycle performance was very poor due to the large
volume change of the reaction Sn ꢀ Li4.4Sn. In this work, the pla-
teau of the reaction of LiSn ꢀ Li7/3Sn is focused. The reaction in-
dicates a long theoretical plateau because the reaction has the largest
lithium-storage reaction of 4/3 lithium ͑301 mAh/g͒ per unit in the
seven reactions as shown in Eq. 1-4. If the lithium-storage reaction
National Institute of Advanced Industrial Science and Technology as-
sisted in meeting the publication costs of this article.
is not Sn ꢀ Li4 Sn but Sn ꢀ Li7/3Sn, the volume change is greatly
.4
decreased. In fact, the charge/discharge cycling curves performed at
+
1
A/g over the voltage range of 0.3–0.7 V vs a Li/Li reference
References
electrode indicate nearly perfect cycle performance as shown in Fig.
1
. P. Poizot, S. Laruelle, S. Grugeon, L. Dupont, and J. M. Tarascon, Nature (Lon-
don), 407, 496 ͑2000͒.
2. X. Wang, X. Chen, L. Gao, H. Zheng, Z. Zhang, and Y. Qian, J. Phys. Chem. B,
08, 16401 ͑2004͒.
. B. Gao, S. Sinha, L. Fleming, and O. Zhou, Adv. Mater. (Weinheim, Ger.), 13, 816
2001͒.
4. W. J. Weydanz, M. Wohlfahrt-Mehrens, and R. A. Huggins, J. Power Sources,
81–82, 237 ͑1999͒.
5. Y. Idota, T. Kubota, A. Matsufuji, Y. Maekawa, and T. Miyasaka, Science, 276,
5
a because the reaction of LiSn ꢀ Li7/3Sn is only occurred in the
windows of 0.3–0.7 V judging from Fig. 2a. In the case of the
high-rate condition, the plateau of LiSn ꢀ Li7/3Sn at 0.4 V in the
discharge process in the case of low rate ͑0.1 A/g͒ is decreased to
1
3
͑
0
.15 V due to the overpotential. It is considered that other lithium
storage potentials of Li7/3Sn → Li Sn → Li Sn → Li3.5Sn
2.5
2.6
+
→
Li4.4Sn are also decreased to less than 0.0 V vs Li/Li resulting
1
395 ͑1997͒.
from overpotential at the high-rate condition. The capacity at the
low-rate condition of 0.1 A/g after the cycling of 10 times at high-
rate condition of 20 A/g is larger than that of 20 A/g after 10 cy-
cling and similar to the that of 0.1 A/g as shown in Fig. 2a. There-
fore, the amount of stored lithium per unit is not 4.4 but near 7/3.
This small volume change based on the storage of 7/3 lithium
caused the good cycle performance. When the potential window is
6
. J. Yang, Y. Takeda, N. Imanishi, and O. Yamamoto, J. Electrochem. Soc., 146,
4009 ͑1999͒.
. M. Winter and J. Besenhard, Electrochim. Acta, 45, 31 ͑1999͒.
. N. Li and C. R. Martin, J. Electrochem. Soc., 148, A164 ͑2001͒.
. N. Tamura, R. Ohshita, M. Fujimoto, S. Fujitani, M. Kamino, and I. Yonezu,
J. Power Sources, 107, 48 ͑2002͒.
10. S. D. Beattie, T. Hatchard, A. Bonakdarpour, K. C. Hewitt, and J. R. Dahn,
J. Electrochem. Soc., 150, A701 ͑2003͒.
11. E. Hosono, S. Fujihara, I. Honma, and H. Zhou, Electrochem. Commun., 8, 284
7
8
9
+
more decreased such as 0.9–0.14 V vs a Li/Li reference electrode
͑2006͒.
at the high-rate condition of 20 A/g, the charge/discharge cycling
curves indicate good cycle performance as shown in Fig. 5b.
In Table I, the specific capacity in the limited voltage regions
among this work and other reported works for high-rate lithium-ion
12. E. Hosono, S. Fujihara, I. Honma, and H. Zhou, J. Electrochem. Soc., 153, A1273
͑2006͒.
1
3. T. Kudo, Y. Ikeda, T. Watanabe, M. Hibino, M. Miyayama, H. Abe, and K. Kajita,
Solid State Ionics, 152–153, 833 ͑2002͒.
4. I. Moriguchi, R. Hidaka, H. Yamada, T. Kudo, H. Murakami, and N. Nakashima,
Adv. Mater. (Weinheim, Ger.), 18, 69 ͑2006͒.
1
storage devices are compared. Other works have not shown the flat
plateau. Hence, the CЈ/⌬V of 2370 mAh/g−1
V
−1
in our work is
15. H. Zhou, D. Li, M. Hibino, and I. Honma, Angew. Chem., Int. Ed., 44, 797 ͑2005͒.
1
1
1
6. J. Wang, I. D. Raistrick, and R. A. Huggins, J. Electrochem. Soc., 133, 457 ͑1986͒.
7. Z. W. Fu and Q. Z. Qin, J. Phys. Chem. B, 104, 5505 ͑2000͒.
8. H. C. Shin, J. Dong, and M. Liu, Adv. Mater. (Weinheim, Ger.), 15, 1610 ͑2003͒.
much larger than those of other works and the range of voltage
change ⌬V ͑0.14 V͒ is very small. The CЈ/⌬V of about
−
1
−1
8
1000 mAh/g
V
of SnO nanofiber is also relatively high. How-
19. H. C. Shin and M. Liu, Adv. Funct. Mater., 15, 582 ͑2005͒.
2
Downloaded on 2015-02-11 to IP 128.122.253.228 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).