Advanced structures in electrodeposited tin base negative electrodes for lithium secondary batteries

Article abstract of DOI:10.1149/1.1568108

Tin anodes deposited electrochemically on a copper foil current collector are studied to develop a next-generation lithium-ion battery with higher energy density. Better cycle performance through ten initial cycles under full charge and discharge conditions was attained by annealing tin electrodeposited on a rough surface copper foil. The annealing process was found to change the main active material from Sn to Cu₆Sn₅ with some minor compounds. Furthermore, a microcolumnar structure of the active material portion was found to be self-organized in accordance with the surface profile of the foil during the first charge-discharge cycle. Advantages of these structural features are discussed in terms of the initial charge and discharge performance, including specific capacity and coulombic efficiency measured by using a three-electrode cell.

Full text of DOI:10.1149/1.1568108

Journal of The Electrochemical Society, 150 ͑6͒ A679-A683 ͑2003͒
A679
0
013-4651/2003/150͑6͒/A679/5/$7.00 © The Electrochemical Society, Inc.
Advanced Structures in Electrodeposited Tin Base Negative
Electrodes for Lithium Secondary Batteries
z
Noriyuki Tamura, Ryuji Ohshita, Masahisa Fujimoto, Maruo Kamino,
and Shin Fujitani
Sanyo Electric Company, Limited, Energy Research and Development Center, Soft Energy Company,
Nishi-ku, Kobe, Hyogo 651-2242, Japan
Tin anodes deposited electrochemically on a copper foil current collector are studied to develop a next-generation lithium-ion
battery with higher energy density. Better cycle performance through ten initial cycles under full charge and discharge conditions
was attained by annealing tin electrodeposited on a rough surface copper foil. The annealing process was found to change the main
active material from Sn to Cu Sn with some minor compounds. Furthermore, a microcolumnar structure of the active material
6
5
portion was found to be self-organized in accordance with the surface profile of the foil during the first charge-discharge cycle.
Advantages of these structural features are discussed in terms of the initial charge and discharge performance, including specific
capacity and coulombic efficiency measured by using a three-electrode cell.
©
2003 The Electrochemical Society. ͓DOI: 10.1149/1.1568108͔ All rights reserved.
Manuscript submitted March 25, 2002; revised manuscript received December 12, 2002. Available electronically April 11, 2003.
The demand for lithium-ion batteries as a power supply for por-
table electric devices has been steadily increasing. Particularly, the
requirement for raising their volume-based energy density is grow-
ing stronger. Graphite is now widely used for the anode of commer-
cially available lithium-ion batteries, but the capacity has already
factors and an advanced electrode structure for better capacity and
cycle performance of lithium alloy electrodes, including those con-
sisting of already studied active materials, are proposed.
Experimental
approached the theoretical limit of C Li.
Four electrodeposited tin electrodes were prepared and exam-
ined, namely, nonannealed/annealed tin electrode with a flat/rough
surface copper foil as current collector. The active material was
electrodeposited on the copper foil with either flat or rough surface
by using the bath described in Table I after surface treatment for the
6
At this background, other group IV elements like tin and silicon
have been gathering considerable attention from many researchers
and are summarized in some reviews,1 because they are expected
to react with far more lithium than carbonaceous materials like
graphite. Figure 1 shows the specific capacity of group IV elements
and Al, calculated on the theoretical base of stoichiometry for inter-
,2
2
1
foil by H SO . The additive agent listed in Table I is a common
2
4
commercial one to reduce Sn whisker generation and prevent SnSO4
from oxidation. The average thickness of the tin layer and the cop-
per foil were 2 and 18 ␮m, respectively. Annealing was conducted at
200°C for 24 h in vacuum atmosphere after the tin electrodeposition.
The mass of the electrodeposited tin was determined by dissolving
the electrode into nitrohydrochloric acid and measuring the Sn con-
centration of its dilute solution by inductively coupled plasma ͑ICP͒
spectrometry. The arithmetic mean roughness measured by a laser
microscopy ͑OLS1100, Olympus Optical Co., Ltd.͒ was 1.19 ␮m for
the rough surface foil and 0.04 ␮m for the flat surface foil, respec-
tively.
3
metallic compounds with Li in the binary alloy phase diagrams.
Providing that an anode from those chemistries in Fig. 1 is com-
bined with a lithium compound like LiCoO as cathode for the cur-
rent lithium-ion battery system, the nonlithiated state should be used
to calculate the gravimetric capacity while the lithiated state should
be applied to the volumetric capacity. The same results are presented
by Winter and Besenhard.1
2
From this point of view, tin-based materials have been studied
for years.4
-19
Tin oxides offer large capacity and good cycle
However, large irreversible capacity at the first
5
-10
performance.
cycle is caused by reduction of tin oxides and formation of lithium
oxide, and still stands in the way of their commercialization. Some
tin-based alloy electrodes of slurry-coating structure onto a current
The electrochemical measurements were conducted by using a
three-electrode test cell. The test electrode size was 20 ϫ 20 mm.
To remove water, the test electrodes were dried at 105°C for 2.5 h
before the measurement. Lithium metal was used as both the counter
and reference electrodes. The electrolyte was ethylene carbonate/
dimethyl carbonate (EC/DMC) ϭ 1/1 (v/v) containing 1 mol dm
LiPF6 .
5,11-17
collector have also been investigated.
In spite of many efforts,
such as employing conductive additives, fine tin particles, and alloy-
ing, to overcome the initial irreversible capacity and poor cycle
performance due to large volume change causing conductivity deg-
radation, reversible capacity is limited to about half the theoretical
Ϫ3
Ϫ1
1,18,19
value of Li4.4Sn, 994 mAh g . Besenhard et al.
showed a
different way, namely, electrodeposited tin anode, suggesting that
very fine tin particles improve cyclability, though the available ca-
pacity still remained less than the theoretical value of Li4.4Sn.
Our goal is to realize improved performance in cyclability and
initial irreversible capacity under full charge and discharge condi-
tions, enabling tin-based negative electrodes to work in nearly the
same volumetric capacity as lithium metal. We have already shown
some encouraging results2
0-22
for the initial cycle performance of the
electrodeposited tin alloy electrodes with nearly theoretical capacity
and durability against repeated charge and discharge cycles.
This work discusses effects of an annealing process after tin
electrodeposition and surface roughness of the copper foil on charge
and discharge performance of electrodeposited tin electrodes in
terms of chemical and physical change of the tin layer. Some crucial
Figure 1. The specific capacity of group IV elements and Al theoretically
calculated in ͑*͒ a charged, or lithiated state, and ͑**͒ discharged, or non-
lithiated state. ͑
͒ Based on the nonlithiated state, the theoretical
z
E-mail: tamura15@sm.energy.sanyo.co.jp
gravimetric capacity of lithium is infinite.
Downloaded on 2014-10-27 to IP 134.245.224.26 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A680
Journal of The Electrochemical Society, 150 ͑6͒ A679-A683 ͑2003͒
Table I. Composition of the electrodeposition bath for tin active
Table II. Characteristics in the first cycle and cycle performance
of the electrodeposited electrodes.
a
material.
Materials
SnSO4
Concentration
40 g dmϪ3
Initial
Capacity ratio
discharge
capacity Initial coulomb
retained after
the 10th
Ϫ3
H SO ͑98%͒
150 g dm
Electrode properties
͑mAh/g͒
efficiency ͑%͒
cycle ͑%͒
2
4
3
Ϫ3
Formalin
Additive agent
5 cm dm
b
3
Ϫ3
Heat-treatment Cu foil
surface
40 cm dm
a
Current density: 0.01 A cmϪ2
.
b
Nonannealed
Flat
892
940
853
905
88
93
85
90
Ͻ38 ͑2nd/1st͒
Produced by Uyemura Kogyo Co., Ltd.
͑
As-deposited͒ Rough
21
33
91
Annealed
Flat
Rough
ϩ
The test cell was fully charged to 0 V vs. Li/Li in three succes-
Ϫ2
first charge and discharge and showed no plateau at either 0.4 or 0.6
V vs. Li/Li in the first discharge because the active material chem-
sive steps of decreasing current density starting at 0.25 mA cm , if
Ϫ2
ϩ
necessary, followed by 0.13 and 0.05 mA cm , and was fully dis-
ϩ
istry changed by the annealing, as discussed here. However, the
surface roughness of the copper foil did not influence the charge and
discharge profiles of both the annealed and nonannealed electrodes.
This result suggests that the active material chemistry is not affected
by the surface roughness of the foil.
Figure 3 shows the cycle performance of the four electrodes. The
annealed electrodes exhibited better cycle performance than those
nonannealed, as shown in Table II. Particularly, the annealed elec-
trode with the rough surface foil retained 91% of its initial capacity
for ten cycles under the full charge and discharge condition, and its
average coulombic efficiency was 98% between the second and
tenth cycle. The annealed electrode with the flat surface foil also
showed good cyclability in the first three cycles; however, the ca-
pacity decreased gradually after that. Therefore, the surface rough-
ness of the foil also affects the cyclability of the annealed electrodes
in a different way from improving the active material chemistry.
charged to 2.0 V vs. Li/Li in the same steps.
The structure of the active material in these electrodeposited
electrodes before and after the first cycle was investigated by scan-
ning electron microscopy ͑SEM, Hitachi S-4500͒ and X-ray diffrac-
tometry ͑XRD, Rigaku Rotaflex, Cu K␣, 40 kV-100 mA͒. The
sample electrodes were buried in epoxy resin and then polished to
observe the SEM images of their cross sections. The lithium-free
electrode samples for XRD were prepared by fixing the electrode on
a glass plate for the lithium-free sample, while the lithium-
containing electrodes were sealed by a polyethylene film.
Results and Discussion
Charge and discharge characteristics.—Charge and discharge
curves of the four electrodes in the first cycle are shown in Fig. 2,
Ϫ2
where the discharge was carried out only at 0.25 mA cm . Almost
no additional discharge capacity was observed in the following cur-
rent density steps. Their initial electrochemical characteristics are
summarized in Table II.
Annealing effects on the active material chemistry.—Both the
nonannealed and annealed electrode with rough surface foil are the
same as the one investigated in Ref. 21. The phase structure of the
nonannealed electrode was comprised of a Cu foil and Sn layer with
a Cu Sn layer thinner than 0.5 ␮m between them. By contrast, that
All the electrodes showed discharge capacity of 850-940 mAh
gϪ1-Sn, which is 2.3-2.5 times as large as the theoretical capacity of
6
5
Ϫ1
graphite for C Li, 372 mAh g . They also showed initial coulom-
of the as-annealed electrode was comprised of the Cu foil,
6
bic efficiency of around 90%, which is comparable to that of graph-
ite commercially employed for current lithium-ion batteries.
The electrode potential of the annealed electrodes, in comparison
with the nonannealed ones, shifted to more negative direction for the
Cu Sn-like phase of 2 ␮m, Cu Sn of 4 ␮m, and Sn layer thinner
than 0.5 ␮m at the top. The flat surface foil provided the same
structures.
3
6
5
Reversible phase structural change of the Cu Sn in the annealed
6
5
electrodes by charge and discharge was observed from the XRD
Figure 2. Initial charge-discharge curves of the electrodeposited electrodes
Figure 3. Cycle performance of the electrodeposited electrodes at 25°C.
ϩ
ϩ
at 25°C. Electrode potential range 0-2.0 V vs. Li/Li ; charge-discharge cur-
Electrode potential range 0-2.0 V vs. Li/Li ; charge-discharge current den-
Ϫ2
Ϫ2
rent density 0.25-0.13-0.05 mA cm . ͑a, • • • •͒. Nonannealed with a flat
sity 0.25-0.13-0.05 mA cm . ͑a, • • • •͒ nonannealed with a flat surface
surface foil, ͑b,
͒ nonannealed with a rough surface foil, ͑c,
foil, ͑b,
͒ nonannealed with a rough surface foil, ͑c,
•
•
•
͒
•
•
•
͒ annealed with a flat surface foil, and ͑d,
rough surface foil.
͒ annealed with a
annealed with a flat surface foil, and ͑d,
͒ annealed with a rough surface
foil.
Downloaded on 2014-10-27 to IP 134.245.224.26 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, 150 ͑6͒ A679-A683 ͑2003͒
A681
The Cu Sn phase obtained by annealing, which results in the
6
5
change of the active material from Sn, is one dominant chemical
factor for improving cyclability of the electrodeposited Sn elec-
trodes. Less swelling of Cu Sn than that of Sn during charge is
6
5
2
3-26
reported to suppress its pulverization.
lithium to Li4.4Sn through Li CuSn, then the chemical reaction
Cu Sn5 reacts with
6
2
5
2
equation is written as follows: Cu Sn ϩ 22Li → Li Sn ϩ 6Cu.
6
5
22
5
The volume change from Cu Sn to the sum of Li Sn and 6Cu is
6
5
22
5
1
calculated to be 260% in this reaction, while Sn swells up by 360%
in the reaction with lithium to Li4.4Sn.
Cu Sn-like phase newly formed by annealing may also improve
3
cyclability, because the annealed electrodes offered better cycle per-
2
5,26
formance than powdered Cu Sn electrodes
under almost the
6
5
same charge and discharge conditions. The existence of Cu Sn-like
3
phase between the Cu Sn layer and the foil enhances the foil to
6
5
stick to the Cu Sn layer. It is considered that the initial capacity
6
5
loss related to Cu Sn-like phase is small. Cu Sn can react with
3
3
26
lithium with large initial irreversible capacity; however, the
Cu Sn layer protects Cu Sn-like phase from reacting with lithium,
6
5
3
as shown in Fig. 4. Also, because Cu Sn-like phase is formed far-
3
thest away from the interface having existed between Sn and Cu
before annealing by diffusion of Sn into Cu foil, the density of Sn in
the Cu Sn-like phase may be less than in the single-crystal Cu Sn.
3
3
Then the ratio of the amount of Sn in this phase to all that in the
annealed electrodes is small.
There was little difference in theoretical volume base energy
density in the lithiated stage between the annealed and nonannealed
electrodes, comprised of Cu Sn and Sn as the active materials,
6
5
respectively. In this work, the thickness of the as-annealed electrode
of 20 ␮m was the same as the nonannealed electrode, where the
Cu Sn was 4 ␮m thick, which should have been 3.1 ␮m if all Sn
6
5
had changed by annealing. This enlargement in thickness was
2
1
caused by the pores in its layer. Providing the volume of both
Cu Sn and Sn changes only in the thickness direction during
6
5
charge, the calculated thickness of the annealed electrode changes
by charge from 20 ͑the Cu Sn of 4 ␮m and the inert part of 16 ␮m͒
6
5
to 26.4 ␮m (Cu Sn → Li Sn ϩ 6Cu of 10.4 ␮m and the inert
6
5
22
5
part of 16 ␮m͒, while that of the nonannealed electrode changes by
charge from 20 ͑the Sn of 2 ␮m and the inert part of 18 ␮m͒ to 25.2
␮
m (Sn → Li Sn of 7.2 ␮m and the inert part of 18 ␮m͒. There-
2
2
5
Figure 4. XRD patterns of the annealed electrodes for ͑a͒ as-annealed, ͑b͒
fore, the advantage of electrodeposited tin electrodes in the volume
base energy density is expected to remain almost the same after
annealing.
ϩ
after the initial charge to 0 V vs. Li/Li , and ͑c͒ after the initial charge and
ϩ
discharge to 2.0 V vs. Li/Li .
Surface roughness effects on the structure of the active ma-
terial.—The effect of the surface roughness of the foil on cyclability
was investigated in terms of the electrode structure after the first
cycle. Figure 5 shows the top and side views of the annealed elec-
trode with the rough surface copper foil after the first cycle. From
patterns as shown in Fig. 4, indicating that the main active material
of the annealed electrodes was the Cu Sn . By contrast, the
6
5
Cu Sn-like phase, which remained the same through the first charge-
3
discharge, is considered to be almost inert to lithium.
Figure 5. SEM images of ͑a͒ top and
͑
b͒ side views of the active material of
the annealed electrode after the initial
ϩ
discharge to 2.0 V vs. Li/Li ͑x2000͒.
Downloaded on 2014-10-27 to IP 134.245.224.26 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A682
Journal of The Electrochemical Society, 150 ͑6͒ A679-A683 ͑2003͒
Figure 6. SEM cross-sectional images
backscattered electron images͒ of the
͑
annealed electrodes after the initial dis-
ϩ
charge to 2.0 V vs. Li/Li ͑a͒ with rough
surface copper foil and ͑b͒ with flat sur-
face copper foil ͑x2500͒.
the top view, the active material is observed to divide into islands by
cracks. From the side view, each island looks like a column. The
cross-sectional dimensions of these microcolumns, i.e., dimensions
lumnar structure cannot be stably formed. In this case, the stress
during charge cannot be released enough to prevent the active ma-
terial from pulverizing and delaminating from the foil. Similar is-
2
1,27
of the islands, are about 10 ␮m .
lands and cracks were observed in previous works.
They also
Figure 6 shows backscattered electron images of cross sections
of the annealed electrodes with rough and flat surface foils after the
first cycle. For the rough surface foil, it is observed that the active
material cracked in accordance with the surface profile of the foil
and formed the microcolumnar structure. For the flat surface foil, by
contrast, the active material did not form this structure and was
likely to be delaminated from the foil.
A black layer whose dark contrast corresponds to existing light
atoms, such as carbon, is observed at the top of the active material in
both electrodes in Fig. 6. This layer may consist of products formed
by reaction of the electrode with electrolyte, which probably caused
the loss of coulombic efficiency.
seemed not to be controlled by the surface profile of the current
collector foil, then the microcolumnar structure was not formed ef-
fectively and stably.
In this way, the idea of the microcolumnar structure can be ex-
tended as a crucial factor for the better cycle performance of lithium
alloy electrodes. The cyclability of the nonannealed electrode with
rough surface foil is poor, probably because the active material Sn is
too brittle or too ductile to form the microcolumnar structure. In
other words, the Cu Sn phase is moderately ductile for forming
6
5
cracks in accordance with the surface profile of the foil and for
maintaining the microcolumnar structure in successive charge-
discharge cycles.
Figure 7 shows schematic models of the structures of the an-
nealed electrodes with flat and rough surface foils, based on this
discussion. It is considered that the effect of the surface roughness
of the foil on improving cyclability results from the formation of the
microcolumnar structure. The active material on the rough surface
foil cracks periodically in accordance with the surface profile of the
foil, in this case, every 10 ␮m, followed by forming the stable
microcolumnar structure. This provides moderate size of the col-
umns and moderate spaces for the active material to swell in either
upper or side directions during charge so that it can strongly stick to
the foil under much reduced stress and strain caused by its volume
change. Thus, the wavy surface profile of the foil is a critical factor
to self-organize the microcolumnar structure of the active material
by charge and discharge.
Conclusion
Almost theoretical capacity based on Sn/Li4.4Sn was found to be
available for an electrodeposited and annealed tin electrode on and
with rough surface copper foil as current collector. Better cycle per-
formance through ten cycles without large sacrifice of their volumet-
ric capacity was also attained. Formation of an advanced structure in
this electrode results in the following improvements:
1
. The active material Cu Sn formed by annealing swells less
6 5
than Sn during charge and is less pulverized.
. Cu Sn-like phase also formed by annealing between the
Cu Sn layer and the foil enhances the foil to stick to the Cu Sn
2
3
6
5
6
5
layer and suppresses delamination of Cu Sn layer from the foil.
6
5
3
. The microcolumnar structure of the active material layer
Because the flat surface foil does not have periodical surface
profile, the dimension of the active material islands divided by
formed on the copper foil with rough surface during the first charge-
discharge cycle accommodates swelled active material in the lithi-
ated stage with much reduced internal stress and strain, and then
pulverizes the active material less and sticks it to the foil.
2
cracks seems to be random and around 20 ␮m , and the microco-
To form and maintain the microcolumnar structure in successive
charge-discharge cycles, tin needs to be alloyed to be moderately
ductile in mechanical property, in this case, forming Cu Sn by the
6
5
annealing process. For better cycle performance, ductility of the
active material, surface profile of the copper foil, the interface layer
between them, and the electrolyte chemistry must be investigated
further.
Sanyo Electric Company, Limited, assisted in meeting the publication
costs of this article.
References
1
2
3
. M. Winter and J. O. Besenhard, Electrochim. Acta, 45, 31 ͑1999͒.
. R. A. Huggins, J. Power Sources, 81–82, 13 ͑1999͒.
. Desk Handbook: Phase Diagrams for Binary Alloys, H. Okamoto, Editor, Ameri-
can Society for Materials, Metals Park, OH ͑2000͒.
4
. B. A. Boukamp, G. C. Lesh, and R. A. Huggins, J. Electrochem. Soc., 128, 725
͑
1981͒.
5
. I. A. Courtney and J. R. Dahn, J. Electrochem. Soc., 144, 2045 ͑1997͒.
Figure 7. Schematic model of the annealed electrode structure with rough
and flat surface copper foil.
6. Y. Idota, T. Kubota, A. Matsufuji, Y. Maekawa, and T. Miyasaka, Science, 276,
1395 ͑1997͒.
Downloaded on 2014-10-27 to IP 134.245.224.26 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, 150 ͑6͒ A679-A683 ͑2003͒
A683
7
. M. Nagayama, T. Morita, H. Ikuta, M. Wakihara, M. Takano, and S. Kawasaki,
18. J. Yang, M. Winter, and J. O. Besenhard, Solid State Ionics, 90, 281 ͑1996͒.
Solid State Ionics, 106, 33 ͑1998͒.
. R. Retoux, T. Brousse, and D. M. Schleich, J. Electrochem. Soc., 146, 2472 ͑1999͒.
. H. Morimoto, M. Nakai, M. Tatsumisago, and T. Minami, J. Electrochem. Soc.,
19. J. O. Besenhard, J. Yang, and M. Winter, J. Power Sources, 68, 87 ͑1997͒.
20. N. Tamura, R. Ohshita, M. Fujimoto, S. Fujitani, M. Kamino, and I. Yonezu, The
41st Battery Symposium in Japan, Nagoya, p. 540 ͑2000͒.
8
9
1
46, 3970 ͑1999͒.
0. N. Li and C. R. Martin, J. Electrochem. Soc., 148, A164 ͑2001͒.
1. O. Mao, R. A. Dunlap, I. A. Courtney, and J. R. Dahn, J. Electrochem. Soc., 145,
195 ͑1998͒.
2. J. Yang, Y. Takeda, N. Imanishi, T. Ichikawa, and O. Yamamoto, J. Power Sources,
9, 220 ͑1999͒.
21. N. Tamura, R. Ohshita, M. Fujimoto, S. Fujitani, M. Kamino, and I. Yonezu, J.
Power Sources, 107, 48 ͑2002͒.
1
1
1
1
1
1
22. N. Tamura, R. Ohshita, M. Fujimoto, S. Fujitani, M. Kamino, and I. Yonezu,
Abstract 251, The Electrochemical Society and the International Society of Elec-
trochemistry Meeting Abstracts, Vol. 2001-2, San Francisco, CA, Sept. 2-7, 2001.
23. K. D. Kepler, J. T. Vaughey, and M. M. Thackeray, Electrochem. Solid-State Lett.,
2, 307 ͑1999͒.
24. M. M. Thackeray, J. T. Vaughey, A. J. Kahaian, K. D. Kepler, and R. Benedek,
Electrochem. Commun., 1, 111 ͑1999͒.
25. D. Larcher, L. Y. Beaulieu, D. D. MacNeil, and J. R. Dahn, J. Electrochem. Soc.,
147, 1658 ͑2000͒.
26. Y. Xia, T. Sakai, T. Fujieda, M. Wada, and H. Yoshinaga, J. Electrochem. Soc., 148,
A471 ͑2001͒.
27. L. Y. Beaulieu, K. W. Eberman, R. L. Turner, L. J. Krause, and J. R. Dahn, Elec-
trochem. Solid-State Lett., 4, A137 ͑2001͒.
4
7
3. G. M. Ehrlich, C. Durand, X. Chen, T. A. Hugener, F. Spiess, and S. L. Suib, J.
Electrochem. Soc., 147, 886 ͑2000͒.
4. H. Sakaguchi, H. Maeta, M. Kubota, H. Honda, and T. Esaka, Electrochemistry
(
Tokyo, Jpn.), 68, 632 ͑2000͒.
5. D. Larcher, L. Y. Beaulieu, O. Mao, A. E. George, and J. R. Dahn, J. Electrochem.
Soc., 147, 1703 ͑2000͒.
6. C. Wang, A. J. Appleby, and F. E. Little, J. Power Sources, 93, 174 ͑2001͒.
7. I. Rom, M. Wachtler, I. Papst, M. Schmied, J. O. Besenhard, F. Hofer, and M.
Winter, Solid State Ionics, 143, 329 ͑2001͒.
1
1
Downloaded on 2014-10-27 to IP 134.245.224.26 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).