The phase transformations and cycling performance of copper-tin alloy anode materials synthesized by sputtering

Article abstract of DOI:10.1016/j.jallcom.2010.08.148

A sputtering technique was adopted to synthesize Sn-Cu thin film electrodes. Island Sn particles were obtained on the copper foil. Cu ₆Sn₅ was spontaneously generated at the interface between Sn and Cu foil. To further improve the cycling stability, Cu source was introduced to increase the formation of Cu₆Sn₅ and to serve as a buffer during cycling. Moreover, the phase and elemental ratio of Sn and Cu varied in the synthesized electrode by alternately adjusting sputtering time for Sn and Cu. The cell synthesized by sputtering Sn for 5 min and Cu for 9 s alternately exhibited the best cycling stability. The 1st charge capacity of cell was 635 mA hg^-1, and the 1st efficiency was even higher than 97%. The capacity remained higher than 500 mA hg^-1 after 15 cycles. The phase transformation of cell was investigated through voltage profile, CV curve and in situ XRD analysis. The in situ XRD analysis confirmed that Cu ₆Sn₅ could react with lithium directly without the presence of Li₂CuSn during cycling. The reaction mechanism of Cu ₆Sn₅ with lithium during cycling was demonstrated to be an alloying process, and the structure of Cu₆Sn₅ was thus a low-temperature monoclinic phase.

Full text of DOI:10.1016/j.jallcom.2010.08.148

Journal of Alloys and Compounds 509 (2011) 123–127
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jallcom
The phase transformations and cycling performance of copper–tin alloy anode
materials synthesized by sputtering
Yu-Sheng Lin , Jenq-Gong Duh , Hwo-Shuenn Sheub
a
a,∗
a
Department of Materials Science and Engineering, National Tsing Hua University, Hsinchu, Taiwan
National Synchrotron Radiation Research Center, Hsinchu Science-Based Industrial Park, Hsinchu, Taiwan
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 24 July 2010
Received in revised form 25 August 2010
Accepted 28 August 2010
Available online 6 September 2010
A sputtering technique was adopted to synthesize Sn–Cu thin film electrodes. Island Sn particles were
obtained on the copper foil. Cu6Sn5 was spontaneously generated at the interface between Sn and Cu foil.
To further improve the cycling stability, Cu source was introduced to increase the formation of Cu6Sn5
and to serve as a buffer during cycling. Moreover, the phase and elemental ratio of Sn and Cu varied in
the synthesized electrode by alternately adjusting sputtering time for Sn and Cu. The cell synthesized
by sputtering Sn for 5 min and Cu for 9 s alternately exhibited the best cycling stability. The 1st charge
Keywords:
Energy storage materials
Thin films
Crystal structure
Synchrotron radiation
Electrochemical reactions
Phase transitions
−1
capacity of cell was 635 mA hg , and the 1st efficiency was even higher than 97%. The capacity remained
−1
higher than 500 mA hg after 15 cycles. The phase transformation of cell was investigated through volt-
age profile, CV curve and in situ XRD analysis. The in situ XRD analysis confirmed that Cu6Sn5 could react
with lithium directly without the presence of Li2CuSn during cycling. The reaction mechanism of Cu6Sn5
with lithium during cycling was demonstrated to be an alloying process, and the structure of Cu6Sn5 was
thus a low-temperature monoclinic phase.
©
2010 Elsevier B.V. All rights reserved.
1
. Introduction
cycling [7,8]. It is argued that the method of reducing metallic Sn
particle sizes discussed in literature might not be an efficient way
to improve cycling performance.
The increasing demand for higher energy density and power
ꢀ
density of batteries in electronic devices has attracted investigators
to develop new materials for lithium-ion batteries. In commer-
cial lithium-ion batteries, the carbonaceous materials are generally
used as anode materials [1]. However, the theoretical capacity of
graphite was only 372 mA hg . As a result, metallic tin element
was used as the anode material, for its higher theoretical capac-
ity, higher lithium packing density and proper operating voltage
The intermetallic compounds (M M) could improve cycling sta-
bility via absorbing the stress induced by large volume change
ꢀ
[9]. The elements M and M were inactive and active element,
respectively. Cu Sn5 was proven to be one of the promising anode
6
−1
materials due to its high reversibility. In literature, several methods
were adopted to synthesize Cu Sn5 as the anode material, including
6
chemical reduction [10,11], electroplating [12,13], electron-beam
deposition [14], ball-milling [15] and sputtering [16]. In this study,
sputtering was selected to synthesize the Sn–Cu electrode, which
[
2]. Nevertheless, the most critical problems in these materials
were the severe volume expansion and contraction during cycling.
Because of the large volume change in the LiXSn alloys, Sn metal
particles were pulverized to lose the electric contact with elec-
trode eventually [3,4], which limited the cycle life of the anode.
To solve this problem, some researchers were devoted to prepar-
ing nanosized Sn particles to reduce the absolute volume change
during cycling and further to improve the electrochemical property
was beneficial to form Cu Sn5 spontaneously at the interface
6
between Sn and Cu without any further heat treatment. Besides,
adjusting the composition of Sn–Cu alloy could optimize cycling
performance by altering its phase using different sputtering time
of Sn/Cu.
In addition, the reaction mechanism of Cu Sn5 during cycling
6
[
5,6]. However, when Sn particles were reduced to nanosizes, the
could be evaluated by in situ XRD analysis [17,18]. Dahn et al. [18]
agglomeration of Sn would become serious, causing cracks, pulver-
ization and capacity fading in the Li–Sn alloy electrodes after long
demonstrated that the reaction of Cu Sn5 with Li during cycling
6
proceeded in two steps. The discharge plateau around 0.4 V was
Cu Sn5 transformation into Li CuSn. The voltage plateau below
6
2
0
.1 V was due to Cu extrusion from the Li CuSn structure to yield
2
Li4.4Sn. According to the binary phase diagram, Cu Sn5 has two
6
∗ Corresponding author at: Department of Materials Science and Engineering,
National Tsing Hua University, No. 101, Section 2, Kuang-Fu Road, Hsinchu, Taiwan
structures, i.e., high-temperature hexagonal ␩-Cu Sn5 and low-
6
ꢀ
temperature monoclinic ␩ -Cu Sn5. Different structures of Cu Sn5
3
0013. Tel.: +886 3 5712686; fax: +886 3 5712686.
6
6
E-mail address: jgd@mx.nthu.edu.tw (J.-G. Duh).
would normally react with lithium through different paths [15,16].
0
925-8388/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2010.08.148

1
24
Y.-S. Lin et al. / Journal of Alloys and Compounds 509 (2011) 123–127
Hexagonal ␩-Cu Sn5 reacted with lithium by forming the interme-
6
ꢀ
diate phase of Li CuSn, while monoclinic ␩ -Cu Sn5 reacted with
2
6
lithium directly without the presence of Li CuSn. In this study, in
2
situ XRD analysis was further used to evaluate the phase transfor-
mations of Cu Sn5 during cycling and to distinguish the structure
6
of Cu Sn5.
6
2.
Experimental
The copper–tin thin films were prepared by sputtering on copper foil using pure
Sn and Cu targets (99.99%, 2 in. in diameter). The gun power of Sn and Cu was 20 W
and 100 W, respectively. The working gas was 99.99% pure argon, the gas flow was
−
4
2
0 sccm, and the working pressure was maintained at 6.0 × 10 Pa during sputter-
ing. The target-to-substrate distance was fixed at 100 mm. Before loading Cu foil into
the chamber, the Cu foil was cleaned by oxalic acid to remove the oxidation layer of
−6
copper. As the vacuum was less than 4.8 × 10 Pa, bias clean was performed under
5
00 V for 10 min to remove the oxidation layer again. Moreover, both Sn and Cu tar-
gets were pre-sputtered for 2 min to clean the surface of target. Different sputtering
time of Sn and Cu was adjusted alternately under constant gun power to change
the composition of copper–tin alloy. The total sputtering time was 2 h. Moreover, to
reveal the effect of introducing Cu source into Sn-based anode material by sputtering
Sn/Cu alternately, the electrode was also deposited only by Sn target for comparison.
The phases were identified with an X-ray diffractometer (XRD, LabX XRD-6000,
Shimadzu, Japan) operated at 30 kV and 20 mA using Cu K␣ with a wavelength of
Fig. 2. Cycling performance of cell which was only deposited by Sn target on the Cu
foil.
1
.5406 A˚ . The deposited amounts of Sn and Cu on the Cu foil were investigated with
a newly designed liquid nitrogen free energy dispersive X-ray spectroscopy (X-Max
SDD, OXFORD). Morphological observations of Sn–Cu electrodes were observed via
scanning electron microscope (SEM, JSM-7600F, JEOL, Japan).
Since the temperature of electrode was very high during sputter-
ing process, the Sn atoms had enough energy to diffuse into the
Electrochemical tests were performed using two-electrode of 2016 type coin
cells (20 mm in diameter and 1.6 mm in thickness). The cells contained an anode
electrode, metallic lithium, polypropylene separator, and electrolyte of 1.0 M LiPF6 in
EC/DMC (1:2, vol.%) plus 2 vol.% of VC. The cells were assembled in an argon glove box
where both moisture and oxygen content were less than 1 ppm. The cells were cycled
between 0.001 V and 1.5 V vs. Li/Li+ at the rate of 0.1 C during first cycle, and then
at the rate of 0.2 C after prolonged cycling. Cyclic voltammetry (CV) measurements
were carried out with a potentiostat (Model 263A, EG&G,) at a scanning rate of
copper matrix and further formed Cu Sn5 at the interface between
6
Sn and Cu. Therefore, the Cu Sn5 with 300 nm in thickness could
6
be revealed at the interface between Sn and Cu foil.
The cycling performance of cell, which was only deposited
by Sn target, is displayed in Fig. 2. The 1st capacity was around
−
1
970 mA hg , which was close to the theoretical capacity of tin
−
1
−
1
(
991 mA hg ). The 10th capacity decreased to 450 mA hg . Above
−1
0
.05 mV s in potential range of 0–1.5 V.
half capacity of cell was dropped after 10 cycles, due to the large
volume change of Sn during cycling. When the cell was disassem-
bled in an argon glove box, parts of Sn were peeled off and separated
from the electrode. In fact, the cycling performance was not good,
The cell designation of in situ XRD experiments was similar to the regular one for
the electrochemical test. To have in situ X-ray pass through the cathode electrode,
a hole with 4 mm diameter was drilled in the upper cover, bottom cover and Ni
spacer of the 2016 coin cell and followed by sealing with a Kapton film. In situ
X-ray studies were carried out in the transmission mode at 16 keV (ꢀ = 0.775 A˚ ) in
the National Synchrotron Radiation Research Center (NSRRC), Taiwan. The exposure
time was 20 s and XRD spectra were recorded on the Mar 345 image plate detector.
All in situ X-ray diffraction patterns were calibrated using standard sample (Ag + Si)
before further analysis.
possibly owing to the small ratio of Cu Sn5/Sn. In other words,
6
insufficient Cu was not able to buffer the volume change during
cycling. Therefore, it was thought to introduce the Cu atom into the
system to increase the formation of Cu Sn5 and to absorb the large
6
volume change during cycling, which might improve the cycling
stability.
3
. Results and discussion
3
.1. The BEI image and cycling performance of electrode
3.2. The X-ray diffraction patterns, microstructure observations,
and quantitative analysis of samples derived from different
sputtering time of Sn/Cu alternately
deposited by Sn
Fig. 1 reveals the BEI image of as-deposited electrode which was
only deposited by Sn target. From the BEI image, the thickness of
the deposited Sn was around 1 ␮m, and the shape was island struc-
ture through BEI image in the observation of cross-section area.
Fig. 3 presents the X-ray diffraction patterns of samples derived
from different sputtering time of Sn/Cu alternately. The phase of
Cu Sn5 and Sn was identified through JCPDS card 45-1488 and 04-
6
0
673, respectively. In order to distinguish the condition of sample
clearly, the sample 3m9s was designated as that the sputtering time
of Sn and Cu was 3 min and 9 s, respectively. The major phase of
sample 3m9s was Cu Sn5. From the BEI image mentioned above,
6
Cu Sn5 was formed at the interface between Sn and Cu during
6
sputtering process. Therefore, introducing Cu source into the sys-
tem could further increase the formation of Cu Sn5. However, the
6
counts of Cu Sn5 decreased and that of Sn increased with sputter-
6
ing time of Sn. As the sputtering time of Sn was above 5 min, the
phase of pure metallic Sn became dominant.
SEM images of samples derived from different sputtering time of
Sn/Cu alternately is presented in Fig. 4. The morphology of electrode
3
m9s displays island structure on the surface of Cu foil, as shown in
Fig. 4a, in which the particle size was around 1 ␮m. As the sputter-
ing time of Sn was increased, the interconnection of particles was
also increased. As a result, the particle size of electrode 6m9s was
the largest among all conditions, as shown in Fig. 4d.
Fig. 1. BEI image of cross-section of electrode which was only deposited by Sn target
on the Cu foil.

Y.-S. Lin et al. / Journal of Alloys and Compounds 509 (2011) 123–127
125
Table 1
Quantitative analysis by EDS for samples derived from different sputtering time of
Sn and Cu.
Sample designation
Measured composition (at.%)
3
4
5
6
m9s
m9s
m9s
m9s
Sn: 50.9(7)
Sn: 74.3(7)
Sn: 84.0(7)
Sn: 94.2(0)
Cu: 49.0(3)
Cu: 25.6(3)
Cu: 15.9(3)
Cu: 5.8(0)
while Cu decreased with the sputtering time of Sn. The composition
of Sn and Cu in the electrode 4m9s were 74.3(7) at.% and 25.6(3)
at.%, respectively. As the sputtering time of Sn was 5 min, Sn con-
tent increased to 84.0(7) at.% and further raised to 94.2(0) at.% with
a sputtering time of 6 min. It was demonstrated that the Sn/Cu ratio
of electrode could be controlled by adjusting different sputtering
time of Sn and Cu.
3.3. The cycling performance of samples derived from different
sputtering time of Sn/Cu alternately
Fig. 3. XRD patterns of sample derived from different sputtering time of Sn and Cu
alternately.
The cycling performance of samples derived from different sput-
tering time of Sn/Cu alternately is displayed in Fig. 5. As compared
with the cell only deposited by Sn target, the cycling performance
of samples derived from different sputtering time of Sn/Cu alter-
nately was much better. It was demonstrated that the introducing
Cu source into Sn-based anode material by sputtering Sn/Cu alter-
nately improved the cycling stability. The cell 5m9s delivered the
best cycling performance among all the samples. The 1st charge
EDS analysis was carried out to evaluate the composition of the
electrodes derived from different sputtering time of Sn/Cu, and the
results are listed in Table 1. The atomic percentage of Sn and Cu
of electrode 3m9s was 50.9(7) at.% and 49.0(3) at.%, respectively,
which was not equivalent to the atomic ratio of Cu/Sn in Cu Sn5.
Most Cu were transformed into Cu Sn5 owing to the elevated tem-
perature during sputtering. Consequently, the electrode 3m9s was
composed of Cu Sn5 and pure Sn, as shown in Fig. 3. The results
of quantitative analysis indicated that the amount of Sn increased,
6
6
−
1
capacity of cell was around 635 mA hg , and 1st efficiency was
−
1
above 97%. The capacity remained higher than 500 mA hg after
15 cycles. The results were attributed to the morphology of elec-
6
Fig. 4. SEM images of sample derived from different sputtering time of Sn and Cu alternately: (a) 3m9s; (b) 4m9s; (c) 5m9s; (d) 6m9s.

1
26
Y.-S. Lin et al. / Journal of Alloys and Compounds 509 (2011) 123–127
Fig. 5. Cycling performance of sample derived from different sputtering time of Sn
and Cu alternately.
Fig. 6. Initial charge–discharge curve of the Sn–Cu thin film electrode 5m9s cycled
between 0 V and 1.5 V vs. Li/Li .
+
trode and the elemental amount adjusted by sputtering time. The
island structure provided more space to tolerate the volume change
than the layered structure. Besides, introducing Cu into the elec-
trode also enhance the retention. Moreover, existing pure metallic
Sn also increased the capacity, and the cycling performance was
the best. However, as the sputtering time of Sn increased to 6 min,
the cycling performance was poorer than that of 5m9s. Although
the initial capacity could be increased by the largest amount of Sn,
need to be recognized.
In the binary Sn–Cu phase diagram, Cu Sn5 has two types of
structure, i.e., high-temperature hexagonal ␩-Cu Sn5 and low-
temperature monoclinic ␩ -Cu Sn5. Different structures of Cu Sn5
reacted with lithium through distinct paths during cycling [15,16].
Consequently, the reaction mechanism of Cu Sn5 alloy during
cycling could be evaluated in details by in situ XRD pattern, as
shown in Fig. 8. The structural changes of copper–tin alloy recog-
nized by the in situ synchrotron X-ray diffraction technique during
discharge process at 15th cycle is indicated in Fig. 8a. The in situ XRD
patterns revealed sequential phase transformations of copper–tin
6
6
ꢀ
6
6
6
the smaller ratio of Cu Sn5/Sn led to insufficient buffer for the vol-
6
ume change during cycling. Thus, the capacity fading was revealed
after prolonged cycling. The capacity of cell 4m9s was smaller than
that of 5m9s, due to less Sn in the electrode. Finally, as the sput-
tering time of Sn was down to 3 min, the capacity was the smallest
among all. The result could be related to the least amount of Sn in
the electrode, as shown in Table 1.
alloy. The phases were pure Sn and Cu Sn5 at the OCV state. The
6
intensity of Cu Sn5 decreased with lithium insertion. It was demon-
6
strated that Cu Sn5 could participate in the reaction with lithium
6
during cycling. However, the reactions between Cu Sn5 and Li were
not displayed in the voltage profile and CV curve. It might be caused
6
3
.4. The phase transformation identification of electrode through
voltage profile, CV, and in situ XRD
by the overlapping of reaction positions of Cu Sn5 and Sn. When
6
the cell was discharged to 0.62 V, Sn and Cu Sn5 were transformed
6
In order to evaluate the phase transformations during cycling of
sample 5m9s, the initial charge/discharge curve of the as-deposited
Sn–Cu thin film electrode in potential window of 0–1.5 V is shown
in Fig. 6. It was apparent that the electrode exhibited three plateaus
in the discharge process, i.e., 0.65 V, 0.50 V and 0.35 V. These poten-
tial plateaus were close to the formation of Li–Sn phases at room
temperature [2]. There were also three potential plateaus at 0.56 V,
into Li Sn . As the lithium was further inserted into Li Sn , the
2
5
2
5
phase of Li Sn changed to LiSn at 0.42 V. As the voltage was below
5
2
0.37 V, Li4.4Sn alloy showed up with the disappearance of LiSn
0.70 V and 0.77 V, respectively, in the charge process, which might
correspond to the delithiation potentials of Li–Sn phases [2]. In fact,
the plateaus in the voltage profile were similar to those in pure Sn
which was reacted with lithium.
In order to make the phase transformations of electrode 5m9s
clearer to be distinguished, CV curve of cell was also performed.
−
1
Fig. 7 shows the cyclic voltammograms at the rate of 0.5 mV s
of cell 5m9s. From the CV curve, three reduction peaks at 0.65 V,
.50 V and 0.35 V could be observed in the cathodic scan, which
0
was related to the generation of Li–Sn materials [14]. Besides, four
oxidation peaks in the anodic scan could be displayed at 0.50 V,
0.63 V, 0.73 V and 0.80 V, respectively. These were associated with
de-alloying in the Li–Sn materials. From the voltage profile and CV
curve mentioned above, it appeared that characteristic reactions
during cycling were attributed to Sn. Nevertheless, the phase of
electrode 5m9s contained Cu Sn5 and pure metallic Sn, as shown
6
in Fig. 3. Both Cu Sn5 and Sn could react with lithium [2,10–15].
6
Fig. 7. Cyclic voltammogram curve of the Sn–Cu thin film electrode 5m9s cycled
between 0 V and 1.5 V vs. Li/Li with a scan rate of 0.05 mV s
+
−1
.
Therefore, the role and the contribution of Cu Sn5 during cycling
6

Y.-S. Lin et al. / Journal of Alloys and Compounds 509 (2011) 123–127
127
cycling [15,16]. When the Cu Sn5 was high-temperature hexagonal
6
phase, the phase transformations of Cu Sn5 was via the formation
6
of Li CuSn intermediate phase during cycling. The voltage plateaus
2
and peaks were at 0.1 V and 0.4 V. However, as the Cu Sn5 was
6
low-temperature monoclinic phase, Cu Sn5 reacted with lithium
6
directly without the presence of Li CuSn phase. Nevertheless, the
2
phase Li CuSn, which was the feature of high-temperature ␩-
2
Cu Sn5, was not shown during cycling. As a result, the reaction
6
mechanism of copper–tin alloy was an alloying process [16], and
ꢀ
the Cu Sn5 alloy might be considered as the low-temperature ␩ -
6
Cu Sn5.
6
4. Conclusions
Sn–Cu thin film electrodes were synthesized by sputtering.
Island Sn particles were deposited on the Cu foil and Cu Sn5 was
6
formed at the interface between Sn and Cu. Introducing Cu source
into Sn-based anode by sputtering Sn/Cu alternately could enhance
the cycling stability as compared with cell deposited only by Sn.
Adjusting different sputtering time of Sn and Cu on the Cu foil alter-
nately could change the phase and composition of electrode. As
the deposition time of Sn and Cu was 5 min and 9 s, respectively,
the cycling stability was the best. The initial capacity of cell was
−
1
around 635 mA hg , and 1st efficiency was as high as 97%. The
−
1
capacity remained higher than 500 mA hg
after 15 cycles. The
electrochemical lithiation reactions of Cu Sn5 were not observed
6
through voltage profile and CV curve. From the in situ XRD analysis,
the transformation of Cu Sn5 during cycling was visible. The phase
6
of Cu Sn5 transformed into Li–Sn alloy directly during cycling with-
6
out the presence of Li CuSn. Therefore, the reaction mechanism of
2
Cu Sn5 was alloying process, and the Cu Sn5 alloy might be con-
6
6
ꢀ
sidered as the low-temperature ␩ -Cu Sn5.
6
Acknowledgements
The authors are grateful to National Science Council, Taiwan
(No. NSC 96-2221-E-007-093) for the partial financial support. We
also thank the group of Dr. Hwo-Shuenn Sheu for their technical
assistance and the National Synchrotron Radiation Research Center
(NRSSC) for the use of synchrotron X-ray diffraction facilities.
References
[
[
[
1] H. Zheng, K. Jiang, T. Abe, Z. Ogumi, Carbon 44 (2006) 203–210.
2] M. Winter, J.O. Besenhard, Electrochim. Acta 45 (2006) 31–50.
3] I.A. Courtney, J.R. Dahn, J. Electrochem. Soc. 144 (1999) 2045–2052.
Fig. 8. In situ XRD analysis of the Sn–Cu thin film electrode 5m9s cycled between 0 V
and 1.5 V vs. Li/Li for the 15th cycle: (a) discharge process and (b) charge process.
+
[4] L.Y. Beaulieu, K.W. Eberman, R.L. Turner, L.J. Krause, J.R. Dahn, Electrochem.
Solid-State Lett. 4 (2001) A137–A140.
[
5] H. Li, Q. Wang, L. Shi, L. Chen, X. Huang, Chem. Mater. 14 (2002) 103–
08.
1
phase. Once the cell reached 0.001 V, the pattern showed mainly
the Li4.4Sn phase. In the process of lithium insertion, the phase of
[
6] P. Limthongkul, H. Wang, E. Jud, Y. Chiang, J. Electrochem. Soc. 149 (2002)
A1237–A1245.
7] I.A. Courtney, W.R. McKinnon, J.R. Dahn, J. Electrochem. Soc. 146 (1999) 59–68.
8] H. Li, L. Shi, W. Lu, X. Huang, L. Chen, J. Electrochem. Soc. 148 (2001) A915–A922.
9] D.G. Kim, H. Kim, H.-J. Sohn, T. Kang, J. Power Sources 104 (2002) 221–225.
[
[
[
Li CuSn did not appear in the sequence of XRD patterns. Moreover,
2
Fig. 8b exhibits in situ XRD patterns of cell during the subsequent
charge process at 15th cycle. The in situ XRD patterns also showed
continuous phase transformations of copper–tin alloy. When the
voltage was increased to 0.64 V, Li4.4Sn alloy disappeared, and the
phase of LiSn showed up. As the lithium was further extracted from
[10] T. Sarakonsri, T. Apirattanawan, S. Tungprasurt, T. Tunkasiri, J. Mater. Sci. 41
2006) 4749–4754.
11] F. Wang, M. Zhao, X. Song, J. Alloys Compd. 439 (2007) 249–253.
[12] S.D. Beattie, J.R. Dahn, J. Electrochem. Soc. 150 (2003) C457–C460.
[13] J.W. Park, S. Rajendran, H.S. Kwon, J. Power Sources 159 (2006) 1409–1415.
14] R.Z. Hu, Y. Zhang, M. Zhu, Electrochim. Acta 53 (2008) 3377–3385.
15] G.X. Wang, L. Sun, D.H. Bradhurst, S.X. Dou, H.K. Liu, J. Alloys Compd. 299 (2000)
L12–L15.
[16] K.-F. Chiu, K.M. Lin, H.C. Lin, W.Y. Chen, D.T. Shieh, J. Electrochem. Soc. 154
2007) A433–A437.
17] S. Sharma, L. Fransson, E. Sjöstedt, L. Nordström, B. Johansson, K. Edström, J.
Electrochem. Soc. 150 (2003) A330–A334.
[18] D. Larcher, L.Y. Beaulieu, D.D. MacNeil, J.R. Dahn, J. Electrochem. Soc. 147 (2000)
1658–1662.
(
[
[
[
LiSn alloy, Li Sn5 reformed with the disappearance of LiSn phase at
2
0
.77 V. As the voltage was above 0.80 V, pure Sn and Cu Sn5 began
6
to be generated. The phases were Cu Sn5 and pure metallic Sn at
6
(
the fully charged state. These phases were identical with those
at the OCV state. It should be noted that the reactions between
active material and lithium were reversible. In literature, differ-
[
ent structures of Cu Sn5 delivered different reaction paths during
6