View Article Online / Journal Homepage / Table of Contents for this issue

Journal of Dynamic Article Links < C

Materials Chemistry
Cite this: J. Mater. Chem., 2011, 21, 10179
www.rsc.org/materials PAPER
High-capacity lithium insertion materials of lithium nickel manganese oxides
for advanced lithium-ion batteries: toward rechargeable capacity more than
300 mA h g
1
Published on 11 April 2011. Downloaded by University of Leeds on 09/06/2013 06:55:23.

Tsutomu Ohzuku,*a Masatoshi Nagayama,ab Kyoji Tsujia and Kingo Ariyoshia

Received 10th December 2010, Accepted 15th March 2011
DOI: 10.1039/c0jm04325g

Lithium nickel manganese oxides Li[NixLi(1/3
2x/3)Mn(2/3
x/3)]O2 (x ¼ 1/2, 2/7, and 1/5) are prepared
and characterized by XRD and FT-IR, and the samples are examined in non-aqueous lithium cells at
room temperature and 55  C. Among these materials LiNi1/2Mn1/2O2 (x ¼ 1/2) shows the highest
operating voltage and the smallest polarization with a rechargeable capacity of ca. 230 mA h g
1 and
Li[Li1/5Ni1/5Mn3/5]O2 (x ¼ 1/5) shows the lowest operating voltage and the largest polarization with
a rechargeable capacity more than 300 mA h g
1. Extraordinarily large rechargeable capacity of
Li[Li1/5Ni1/5Mn3/5]O2 together with an anomalously long voltage plateau at 4.5 V only observed at first
charging process is examined by window-opening charge and discharge, continuous charge and
discharge combined with differential chronopotentiometry at room temperature and at 55  C, and
possible mechanisms are discussed in terms of lithium insertion scheme.

Introduction To increase the gravimetric capacity of lithium insertion mate-

rials, light elements are better than heavy elements among
During the past 25 years, we have systematically investigated the chemical species listed in the periodic table. Consequently,
positive- and negative-electrode materials for high-energy lithium 3d-transition metal dioxides have been extensively
density lithium batteries under a framework of lithium insertion examined in terms of solid solution or superlattice, leading to
materials. According to a hard sphere model,1 the maximum lithium aluminium cobalt nickel dioxides,8,9 lithium nickel
volumetric capacity is given when a solid-state redox reaction of manganese dioxides with or without cobalt,5 and superstructural
Me3+ and Me4+ concurrent with lithium-ion insertion and lithium nickel manganese dioxide,10 specifically LiAl0.05Co0.15-
extraction, i.e., Ni0.8O2 and LiCo1/3Ni1/3Mn1/3O2 currently used in lithium-ion
batteries. The theoretical gravimetric capacity ranges from
(1) 275 mA h g
1 for LiNiO2 to 310 mA h g
1 for hypothetical LiTiO2,
or, from 295 mA h g
1 for ,NiO2 to 335 mA h g
1 for ,TiO2. The
occurs in a topotactic manner. In eqn (1) (cp) and (oct) indicate rechargeable capacity attained in laboratory cells is around
the close(st) packing of oxygen ions and octahedral sites, 200 mA h g
1 out of ca. 300 mA h g
1 of theoretical capacity.
respectively, , denotes vacant octahedral sites, and Me indi- An introduction of scrap-and-built-type reactions substituting
cates transition metal ions. The specific volumetric capacity is for lithium insertion reactions is an option to break the limitation
calculated using Shannon ionic radii to be 1.56 A h cm
3 for of volumetric and gravimetric capacity. Scrap-and-built-type
,MeO2 or 1.43 A h cm
3 for LiMeO2. No exception can be reactions include lithium metal, lithium metal alloys, and
found in volumetric capacity of lithium insertion materials.2–7 ‘‘conversion’’ electrodes.11,12 These electrodes, however, are
In other words, the volumetric capacity is limited to below usually poor in cycle-ability compared to lithium insertion
1.4 A h cm
3. The gravimetric capacity is also an important electrodes.
factor in evaluating materials for high-energy density batteries. As briefly described above, rechargeable capacity in both
volumetric and gravimetric capacities will saturate soon if
a concept of lithium insertion materials based on eqn (1) is the
a
Department of Applied Chemistry, Graduate School of Engineering, only way to systematic studies on materials for high-energy
Osaka City University (OCU), Sugimoto 3-3-138, Sumiyoshi, Osaka, density lithium-ion batteries. An innovative concept is necessary
558-8585, Japan. E-mail: [email protected]; Tel: to break through such a limitation. To establish a new concept,
+81-6-6605-2693
b
we need anomalous experimental facts that we cannot explain by
Green Energy Business Promotion Office, Energy Company, Panasonic
Corporation, Moriguchi, Osaka, 570-8511, Japan. E-mail: nagayama. the previous concept on lithium insertion materials based on eqn
[email protected]; Tel: +81-6-6998-4552 (1). In this paper, we report a material showing rechargeable

This journal is ª The Royal Society of Chemistry 2011 J. Mater. Chem., 2011, 21, 10179–10188 | 10179
 View Article Online

capacity more than a critical limit of 300 mA h g
1 and discuss 0.15 ml of electrolyte was injected with a syringe. Other sets of
whether or not the material opens new window toward an experimental conditions are given in the Results and discussion
innovation in materials for advanced lithium-ion batteries. section.

Experimental
Lithium nickel manganese oxides were prepared from nickel
Results and discussion
manganese double hydroxides of MX-114 (Ni/Mn ¼ 0.99/1.01 in Characterization of samples
molar ratio; Tanaka Chemical Corp. Ltd., Japan), MX-005-1
(Ni/Mn ¼ 1.00/2.00) and MX-003-1 (Ni/Mn ¼ 1.04/2.96) Fig. 1 shows the XRD patterns of lithium nickel manganese
prepared by a coprecipitation method.13 To prepare a lithium oxides together with Li2MnO3. All XRD patterns of lithium
Published on 11 April 2011. Downloaded by University of Leeds on 09/06/2013 06:55:23.

nickel manganese oxide at Ni/Mn ¼ 1/1 with Li/(Ni + Mn) ¼ 5/4, nickel manganese oxides are similar to those of a-NaFeO2-
the powder with the molar ratio of LiOH/NiMnO3 ¼ 5/2 was well structural type. All main diffraction lines of lithium nickel
mixed with a mortar and pestle by hand. The reaction mixture manganese oxides (a), (b), and (c) in Fig. 1 are indexed in
was pressed into pellet and heated at 1000  C for 12 h in air. hexagonal setting. Although Li2MnO3 (d) is indexed in mono-
NiMnO3 was prepared by heating MX-114 at 600  C for 4 h. To clinic setting, XRD patterns of these four samples are quite
prepare lithium nickel manganese oxides at Ni/Mn ¼ 1/2 or 1/3 similar. Differences are selective line broadening and diffraction
respectively with Li/(Ni + Mn) ¼ 4/3 or 3/2, a reaction mixture of lines observed in 25–35 in 2q. As seen in Fig. 1, selective line
lithium hydroxide and MX-005-1 or MX-003-1 was heated at broadening of (110) line is seen in (a) while a split due to Ka1 and
500  C for 12 h and then at 1000  C for 12 h in air. Heating and Ka2 is seen in (d). Weak diffraction lines at 26 , 27 , 30 , and 35
cooling rates were approximately 5  C min
1. The reaction in 2q are seen in Fig. 1(b) and (c). Miller indexes for these
products were characterized by powder X-ray diffraction using diffraction lines can be given as (1/3,1/3,0), (1/3,1/3,1), (1/3,1/3,2),
an X-ray diffractometer (XRD-6100, Shimadzu, Co. Ltd., and (1/3,1/3,3). The extra diffraction lines are quite similar to
Japan) with FeKa radiation, equipped with a graphite those of (020), (110), (11
1), and (021) in Fig. 1(d) for Li2MnO3
monochromator. Tube voltage and current applied were 40 kV resulting from one-to-two cation ordering of lithium and
and 15 mA, respectively. The XRD data were recorded from 10 manganese ions in a transition metal layer, forming a [O3  O3]
to 104 in 2q at a scan rate of 0.5 min
1. Lattice constants of R30 superlattice based on a basal net in Wood notation.14 The
reaction products were calculated by a least squares method extra diffraction lines in Fig. 1(b) and (c) suggest such a cation
using X-ray diffraction lines. FT-IR spectra of the samples were ordering in the transition metal layer.
obtained by a KBr method by using a Fourier-transform infrared The hexagonal lattice parameters of a-axis, c-axis, and c/a-
spectrometer (FT-IR 8300, Shimadzu, Co. Ltd., Japan). A KBr ratio as a function of x in xLi2MnO3$(1
x)LiNi1/2Mn1/2O2 are
crystal was freshly crushed, grounded and granulated, and then shown in Fig. 2. In drawing Fig. 2, the monoclinic lattice
about 3 mg of the sample was well mixed with ca. 200 mg of the
KBr. The KBr pellet was prepared by pressing the mixture under
vacuum for 5 minutes. FT-IR spectra were recorded in wave-
number between 400 and 2000 cm
1.
An electrochemical cell consists of sample-positive and
lithium-negative electrodes separated by two sheets of porous
membrane (Celgard 2500 or 2320, Hoechst Celanese Co. Ltd.,
USA), which are placed in a cavity of a container. The apparent
electrode area is 2 cm2. A container consists of two stainless steel
plates (5 mm thick) separated by a Teflon spacer (2 mm thick) in
which a ca. 25  35 mm2 window is made. To prepare positive
electrodes, black viscous slurry was cast onto aluminium foil
with a blade, which consisted of 88 wt% sample, 6 wt% acetylene
black, 6 wt% polyvinylidene fluoride dissolved in N-methyl-2-
pyrolidone (NMP). NMP was evaporated at 60  C for 1 h under
vacuum, and finally the electrodes were dried under vacuum at
140  C for at least 12 h. Lithium-negative electrode was prepared
by pressing a lithium foil onto a stainless steel plate. The lithium
foil was freshly rolled on a lithium chip cut out from a lithium
rod (purity 99% Rare Metallic Co. Ltd., Japan). The electrolyte
Fig. 1 XRD patterns of lithium nickel manganese oxides prepared at (a)
used was 1 M LiPF6 dissolved in ethylene carbonate (EC)/
Ni/Mn ¼ 1/1 with Li/(Ni + Mn) ¼ 5/4, (b) Ni/Mn ¼ 1/2 with Li/(Ni +
dimethyl carbonate (DMC) (3/7 by volume) solution (Kishida
Mn) ¼ 4/3, (c) Ni/Mn ¼ 1/3 with Li/(Ni + Mn) ¼ 3/2, and (d) Li2MnO3.
Chemical Co. Ltd., Japan). In fabricating cells, all materials Miller indexes and lattice parameters are given by assuming a hexagonal
except the electrolyte and lithium metal were dried at 40  C for at lattice for (a), (b), and (c), and a monoclinic lattice for (d). Lattice
least 2 h under vacuum to avoid possible contamination with parameters are (a) a ¼ 2.883 A,  c ¼ 14.268 A, (b) a ¼ 2.869 A, c ¼
water. The cells were assembled in an argon-filled glove box.  (c) a ¼ 2.863 A,
14.270 A,  c ¼ 14.267 A,
 and (d) a ¼ 4.930 A,
 b ¼ 8.518 A,

After the electrodes and a separator were placed in the cavity, ca.  and b ¼ 109.28 .
c ¼ 5.021 A,

10180 | J. Mater. Chem., 2011, 21, 10179–10188 This journal is ª The Royal Society of Chemistry 2011
 View Article Online
Published on 11 April 2011. Downloaded by University of Leeds on 09/06/2013 06:55:23.

Fig. 2 Hexagonal lattice parameters of (a) a-axis, (b) c-axis, and (c) c/
a as a function of x in xLi2MnO3$(1
x)LiNi1/2Mn1/2O2. The samples Fig. 3 FT-IR spectra of lithium nickel manganese oxides of (a) LiNi1/2-
are the same as shown in Fig. 1. Open and closed circles given at x ¼ 0 are Mn1/2O2, (b) Li[Li1/7Ni2/7Mn4/7]O2, (c) Li[Li1/5Ni1/5Mn3/5]O2, and (d) Li
LiNi1/2Mn1/2O2 prepared at Ni/Mn ¼ 1/1 with Li/(Ni + Mn) ¼ 1/1 (open [Li1/3Mn2/3]O2 (Li2MnO3).
circles) and 5/4 (closed circles).

parameters of Li2MnO3 are converted to approximate hexagonal Electrochemical charge and discharge at room temperature
lattice parameters using equations

. Fig. 4 shows the charge and discharge curves of LiNi1/2Mn1/2O2,
pffiffiffi
ah ¼ am = 3 þ bm 2 and ch ¼ 3cm sin b; Li[Li1/7Ni2/7Mn4/7]O2 and Li[Li1/5Ni1/5Mn3/5]O2 examined in
lithium cells at room temperature (ca. 25  C). The operating
where am, bm, cm, and b are monoclinic lattice parameters, and, voltage of the first charging curve of LiNi1/2Mn1/2O2 monoto-
ah and ch are the approximate hexagonal lattice parameters. The nously increases from 3.7 to 4.7 V, as shown in Fig. 4(a). The
a-axis linearly decreases from 2.89 to 2.84 A  and c-axis also operating voltages in the first charging curves of Li[Li1/7Ni2/7-

linearly decreases from 14.29 to 14.23 A with increasing x, while Mn4/7]O2 and Li[Li1/5Ni1/5Mn3/5]O2 almost linearly increase up
the c/a value increases. This clearly indicates that these samples to an inflection point at 4.5 V, and then it follows a plateau at
follow a Vegard’s rule. The end member at x ¼ 1 is Li2MnO3 or 4.5 V. At a plateau voltage of 4.5 V in Fig. 4, the charge
alternatively Li[Li1/3Mn2/3]O2 prepared at Ni/Mn ¼ 0 with Li/ capacities of LiNi1/2Mn1/2O2, Li[Li1/7Ni2/7Mn4/7]O2 and Li[Li1/5-
Mn ¼ 2 and that at x ¼ 0 is LiNi1/2Mn1/2O215 prepared at Ni1/5Mn3/5]O2 are observed to be ca. 40, 120, and 150 mA h g
1,
Ni/Mn ¼ 1/1 with Li/(Ni + Mn) ¼ 1/1 or 5/4. Therefore, the respectively. When the Mn content increases, the charge capacity
sample prepared at Ni/Mn ¼ 1/2 with Li/(Ni + Mn) ¼ 4/3 is in voltage below 4.5 V decreases although the capacity at 4.5 V
represented as Li[Li1/7Ni2/7Mn4/7]O2, which is one-to-two increases. For subsequent cycles, all samples exhibit stable cycle
mixture of Li[Li1/3Mn2/3]O2 and LiNi1/2Mn1/2O2, i.e., x ¼1/3 in behaviors as shown in Fig. 4. The rechargeable capacities are
Fig. 2. The sample at Ni/Mn ¼ 1/3 with Li/(Ni + Mn) ¼ 3/2 is the 205 mA h g
1 for LiNi1/2Mn1/2O2, 220 mA h g
1 for Li[Li1/7Ni2/7-
one-to-one mixture (x ¼ 1/2), represented as Li[Li1/5Ni1/5Mn3/5] Mn4/7]O2, and 240 mA h g
1 for Li[Li1/5Ni1/5Mn3/5]O2 at room
O2 hereafter. Sample notation to describe a series of solid solu- temperature. A crossing point between charge and discharge
tion between Li2MnO3 and LiNi1/2Mn1/2O2 will be described in curves for each sample is 95, 80, and 75 mA h g
1, respectively,
later section. In order to characterize the samples, the FT-IR for (a), (b), and (c) in Fig. 4, suggesting the increase in
spectra are measured and shown in Fig. 3. Absorption bands in polarization.
wavenumber ranging from 400 to 800 cm
1 are commonly In order to examine the solid-state redox reactions in lithium
observed among these samples. Fine structures are seen in Fig. 3 nickel manganese oxides, dQ/dE are calculated from numerical
(d) for Li2MnO3 although LiNi1/2Mn1/2O2 shows rounded and data observed in steady-state charge and discharge curves in
ill-defined two absorption bands overlapping at 520 and 575 Fig. 4 and the dQ/dE vs. E plots are illustrated in Fig. 5. LiNi1/2-
cm
1 in Fig. 3(a). IR spectra in Fig. 3(b) and (c) contain Mn1/2O2 shows two well-defined oxidation peaks at 3.77 and
absorption bands at about 445, 530, and 620 cm
1, which may be 4.37 V. The reduction peaks corresponding to oxidation peaks
the sum of the intensities of the absorption bands in Fig. 3(a) and are also observed at 3.74 and 4.34 V, indicating that the reactions
(d) in the molar ratio of one to two for (b) and one to one for (c). are reversible. If we define mid-point voltage between them as

This journal is ª The Royal Society of Chemistry 2011 J. Mater. Chem., 2011, 21, 10179–10188 | 10181
 View Article Online

shape is rounded and ill-defined. Oxidation and reduction peaks

around 3.75 and 4.35 V may be due to the solid-state redox
reactions observed for LiNi1/2Mn1/2O2. The reduction peak at
3.28 V and the oxidation peak at ca. 3.62 V cannot be seen in
Fig. 5(a) for LiNi1/2Mn1/2O2. The redox peak in voltage below
ca. 3.7 V is characteristic of Li[Li1/7Ni2/7Mn4/7]O2 and Li[Li1/5-
Ni1/5Mn3/5]O2.

Electrochemical charge and discharge at 55  C

Published on 11 April 2011. Downloaded by University of Leeds on 09/06/2013 06:55:23.

As described in the previous section, Li[Li1/7Ni2/7Mn4/7]O2 and

Li[Li1/5Ni1/5Mn3/5]O2 show kinetic problems. Polarization
usually decreases when an electrochemical system is heated. In
order to examine the effect of temperature upon the electro-
chemistry of LiNi1/2Mn1/2O2, Li[Li1/7Ni2/7Mn4/7]O2 and Li[Li1/5-
Ni1/5Mn3/5]O2, the cells are operated at 55  C and the results are
compared to those at room temperature.
Fig. 6 shows the charge and discharge curves of the samples
examined in lithium cells at 55  C. Because we need all history
from the first cycle to that approaching steady-state charge and
discharge curves, raw data observed at 55  C are shown in Fig. 6.
At the plateau voltage of 4.5 V, charge capacities are 60 mA h g
1
for LiNi1/2Mn1/2O2, 140 mA h g
1 for Li[Li1/7Ni2/7Mn4/7]O2, and
200 mA h g
1 for Li[Li1/5Ni1/5Mn3/5]O2, which are larger than
Fig. 4 Charge and discharge curves of (a) LiNi1/2Mn1/2O2 examined at those examined at room temperature, as seen in Fig. 4.
0.17 mA cm
2 (12.8 mA g
1), (b) Li[Li1/7Ni2/7Mn4/7]O2 at 0.10 mA cm
2 Rechargeable capacities observed at 55  C are 235 mA h g
1 for
(9.7 mA g
1), and (c) Li[Li1/5Ni1/5Mn3/5]O2 at 0.10 mA cm
2 (8.2 mA g
1)
in non-aqueous cell at room temperature (ca. 25  C). Electrode mix used
is (a) 44.4 mg weight with 129 mm thick, (b) 35.1 mg with 97.6 mm, or (c)
41.5 mg with 104.6 mm. The 1st to 5th cycles are shown in these figures. The
samples are the same as (a) to (c) in Fig. 1 and 3.

Fig. 5 Differential chronopotentiograms of (a) LiNi1/2Mn1/2O2, (b) Li

[Li1/7Ni2/7Mn4/7]O2, and (c) Li[Li1/5Ni1/5Mn3/5]O2 calculated from
numerical data obtained at the 2nd cycle in Fig. 4.

redox potential, the electrochemical reaction of LiNi1/2Mn1/2O2

Fig. 6 Charge and discharge curves of (a) LiNi1/2Mn1/2O2 examined at
consists of two solid-state redox reactions characterized by 3.75 0.17 mA cm
2 (12.3 mA g
1), (b) Li[Li1/7Ni2/7Mn4/7]O2 at 0.10 mA cm
2
and 4.35 V of redox potential. Shapes of oxidation and reduction (7.5 mA g
1), and (c) Li[Li1/5Ni1/5Mn3/5]O2 at 0.10 mA cm
2 (9.3 mA g
1)
peaks in Fig. 5(a) indicate that the reaction proceeds in homo- in non-aqueous cell at 55  C. Electrode mix used is (a) 46.1 mg weight
geneous or one phase. Li[Li1/7Ni2/7Mn4/7]O2 and Li[Li1/5Ni1/5- with 127.8 mm thick, (b) 45.2 mg with 128.2 mm, and (c) 36.6 mg with
Mn3/5]O2 show some oxidation and reduction peaks, but the 102.6 mm. The 1st to 5th cycles are shown in these figures.

10182 | J. Mater. Chem., 2011, 21, 10179–10188 This journal is ª The Royal Society of Chemistry 2011
 View Article Online

LiNi1/2Mn1/2O2, 270 mA h g
1 for Li[Li1/7Ni2/7Mn4/7]O2, and

310 mA h g
1 for Li[Li1/5Ni1/5Mn3/5]O2. The rechargeable
capacities increase by heating especially for Li[Li1/5Ni1/5Mn3/5]
O2. It shows the rechargeable capacity more than 300 mA h g
1.
To show the solid-state redox potential, the dQ/dE as a func-
tion of E is calculated from numerical data observed at the 2nd
cycle in Fig. 6 and shown in Fig. 7. General observations on
solid-state redox potentials in Fig. 7 are the same as described in
Fig. 5 except the shape of peak and its height. At 55  C, all of
oxidation and reduction peaks are shaped up nearly well-defined
peaks compared to those at room temperature in Fig. 5. The
Published on 11 April 2011. Downloaded by University of Leeds on 09/06/2013 06:55:23.

biggest difference between data in Fig. 5 and 7 is the reduction

peak at 3.28 V for Li[Li1/7Ni2/7Mn4/7]O2 and Li[Li1/5Ni1/5Mn3/5]
O2. A peak potential at 3.28 V is insensitive to temperature, but
their values in dQ/dE at 55  C are about two times larger than
those at room temperature. Increase in dQ/dE at 3.28 V is derived
from the increase in discharge capacity in voltage below 3.5 V in
Fig. 6, resulting in lower operating voltage than that of LiNi1/2- Fig. 8 Charge and discharge curves of Li/Li[Li1/5Ni1/5Mn3/5]O2 cell
Mn1/2O2, although the rechargeable capacity of Li[Li1/5Ni1/5- operated at a rate of 0.15 mA cm
2 (8.7 mA g
1) in voltage ranging from
2.0 to 5.0 V at room temperature (ca. 25  C). The electrode mix used is
Mn3/5]O2 is 70 mA h g
1 larger than that of LiNi1/2Mn1/2O2.
39.1 mg weight with 100 mm thick. The 1st to 10th cycles are shown in this
figure.

Electrochemistry of Li[Li1/5Ni1/5Mn3/5]O2
The rechargeable capacity more than 300 mA h g
1 is the highest 260 mA h g
1, drawing 4.5 V plateau. On first discharge, oper-
value reported so far for a series of lithium insertion materials, ating voltage linearly decreases from 4.4 to 3.1 V, and the cell
which is very attractive for advanced lithium-ion batteries. cycles mainly in voltage ranging from 3.0 to 4.5 V. As clearly seen
However, the voltage shape at the first charge is quite different in Fig. 8, the loss of capacity during the first cycle is observed to
from that for subsequent cycles. Consequently, the irreversible be ca. 70 mA h g
1 and the rechargeable capacity is 230 mA h g
1
capacity seems to be large compared to that of LiCoO2. In order at room temperature.
to examine whether or not Li[Li1/5Ni1/5Mn3/5]O2 can be used as In order to examine electrochemical reactions at 4.5 V plateau,
a positive-electrode material for lithium-ion batteries, the irre- the Li/Li[Li1/5Ni1/5Mn3/5]O2 cell is cycled in such a window-
versible loss of capacity during the first cycle and cycling opening condition that charge capacity increases by 25 mA h g
1
behavior during subsequent cycles are studied in detail at room every cycle until the voltage reaches 5.0 V. The result is shown in
temperature and 55  C. Fig. 8 shows the charge and discharge Fig. 9. In capacity ranging from 0 to 110 mA h g
1, discharge
curves of a Li/Li[Li1/5Ni1/5Mn3/5]O2 cell examined at room voltage follows just below charging curves. The irreversible
temperature (ca. 25  C). On the first charge, operating voltage
linearly increases to 4.45 V until charge capacity reaches 115 mA
h g
1 and then stays at 4.5 V until charge capacity reaches

Fig. 9 Window-opening charge and discharge curves of a Li/Li[Li1/5Ni1/5-

Mn3/5]O2 cell operated at a rate of 0.15 mA cm
2 (9.9 mA g
1) in voltage
Fig. 7 Differential chronopotentiograms of (a) LiNi1/2Mn1/2O2, (b) Li ranging from 2.0 to 5.0 V at room temperature (ca. 25  C). The charge
[Li1/7Ni2/7Mn4/7]O2, and (c) Li[Li1/5Ni1/5Mn3/5]O2 calculated from capacity is increased by 25 mA h g
1 every cycle until the voltage reaches
numerical data obtained at the 2nd cycle in Fig. 6 at 55  C. 5 V. The electrode mix used is 34.5 mg weight with 94 mm thick.

This journal is ª The Royal Society of Chemistry 2011 J. Mater. Chem., 2011, 21, 10179–10188 | 10183
 View Article Online

capacity is observed to be ca. 10 mA h g
1 in this range. At 4.5 V

plateau in 110 to 280 mA h g
1, the irreversible capacity increases
cycle by cycle from 10 to 70 mA h g
1 and then it stays at that
value even when the cell cycles in voltage ranging from 2 to 5 V,
i.e., the rechargeable capacity of 250 mA h g
1 and irreversible
capacity of 70 mA h g
1. As seen in Fig. 9, a 4.5 V plateau is
always observed unless the capacity at 4.5 V plateau is consumed.
Fig. 10 shows the result on the window-opening experiment of
Li[Li1/5Ni1/5Mn3/5]O2 at 55  C. General observations are the
same as those at room temperature, except their values. The
irreversible capacity observed at 55  C is 40 mA h g
1, which is
Published on 11 April 2011. Downloaded by University of Leeds on 09/06/2013 06:55:23.

30 mA h g
1 smaller than that at room temperature. The

rechargeable capacity at 55  C is 300 mA h g
1, which is 70 mA
h g
1 larger than that at room temperature. A long plateau at 4.5
V, specifically from 4.43 to 4.54 V in capacity ranging from 120 to
300 mA h g
1 in Fig. 10, is 180 mA h g
1 at 55  C, which is longer
than that at room temperature. As clearly seen in Fig. 10, one has Fig. 11 Charge and discharge curves of a Li/Li[Li1/5Ni1/5Mn3/5]O2 cell
to charge Li[Li1/5Ni1/5Mn3/5]O2 once to the voltage above 4.5 V operated at a rate of 0.15 mA cm
2 (10.1 mA g
1) in voltage ranging from
with respect to Li to have the attractive rechargeable capacity. 2.0 to 5.0 V at 55  C for 20 cycles. The electrode mix used is 33.9 mg
weight with 94 mm thick.

Twenty-cycle test of Li/Li[Li1/5Ni1/5Mn3/5]O2 cell at 55  C The operating voltage on charge decreases by cycling. Charge
and discharge capacities in mA h g
1 together with coulombic
Although an electrochemical reaction at the 4.5 V plateau is efficiency in % as a function of cycle number are shown in Fig. 12.
specifically not known yet, it is interesting to know whether or The capacities decrease from 310 mA h g
1 at the 1st discharge to
not the rechargeable capacity more than 300 mA h g
1 is 300 mA h g
1 at the 6th and then almost linearly increase as
sustainable at 55  C. The cycle test of a Li/Li[Li1/5Ni1/5Mn3/5]O2 a function of cycle number. The charge and discharge coulombic
cell at 55  C is carried out, and the result is shown in Fig. 11. The efficiencies are more than 99% except 90% for the 1st cycle. The
cell is operated at a rate of 0.15 mA cm
2 or 10.1 mA g
1 based on average efficiency is calculated to be 99.75% from the values
the weight of Li[Li1/5Ni1/5Mn3/5]O2 in voltage ranging from 2.0 to observed from the 2nd to 20th cycle. These results suggest that
5.0 V at 55  C. The sample shows the rechargeable capacity more
than 300 mA h g
1 during 20 cycles, as is expected. Almost all
discharge curves converge at two points of 3.85 V with 70 mA h
g
1 and 2.91 V with 295 mA h g
1. Before the first crossing point
in 0–70 mA h g
1, the operating voltage increases by cycling.
Conversely, the operating voltage on discharge between two
crossing points, i.e., 70–295 mA h g
1, decreases. After the
second crossing point, the capacity increases after the 4th cycle.

Fig. 10 Window-opening charge and discharge curves of a Li/Li[Li1/5- Fig. 12 (a) Charge and discharge capacities and (b) coulombic efficiency
Ni1/5Mn3/5]O2 cell operated at a rate of 0.15 mA cm
2 (8.8 mA g
1) in as a function of cycle number observed for a Li/Li[Li1/5Ni1/5Mn3/5]O2 cell
voltage ranging from 2.0 to 5.0 V at 55  C. The charge capacity is at 55  C in Fig. 11. Closed and open circles in (a) are respectively charge
increased by 25 mA h g
1 every cycle until the voltage reaches 5 V. The and discharge capacities. The coulombic efficiency at the first cycle is
electrode mix used is 38.9 mg weight with 119 mm thick. 90%, which is outside the scale in (b).

10184 | J. Mater. Chem., 2011, 21, 10179–10188 This journal is ª The Royal Society of Chemistry 2011
 View Article Online

reactions associated with high rechargeable capacity are still

changing by cycling at 55  C in addition to irreversible change at
4.5 V plateau observed at the 1st charge. In order to examine
whether or not such a change in electrochemical reaction is
involved, the differential chronopotentiograms are calculated
from numerical data obtained at the 2nd, 10th and 20th cycles and
shown in Fig. 13. The differential chronopotentiogram at the 2nd
cycle in Fig. 13(a) is consistent with that shown in Fig. 7, indi-
cating that the data are highly reproducible in spite of the irre-
versible change at the 1st cycle. As clearly seen in Fig. 13, the
reduction peak shifts toward the cathodic direction, specifically
Published on 11 April 2011. Downloaded by University of Leeds on 09/06/2013 06:55:23.

3.28 to 3.02 V via 3.12 V, while dQ/dE values are almost the
same. A striking feature is that the oxidation peaks at 3.04 and
3.18 V develop by cycling in return for the decrease in oxidation
peaks at 3.90 and 4.40 V. These two oxidation peaks at 3.04 and
3.18 V are not observed at all in initial cycles. Because the
reduction peak at ca. 3.02 V and oxidation peaks at 3.04 and
Fig. 14 Charge and discharge curves of a Li/Li[Li1/5Ni1/5Mn3/5]O2 cell
3.18 V seem to relate each other, the redox reactions character-
operated at a rate of 0.15 mA cm
2 (9.9 mA g
1) in voltage ranging from
ized by these reduction and oxidation peaks around 3.1 V may be
2.0 to 5.0 V at 85  C. The electrode mix used is 34.8 mg weight with 98 mm
new redox reactions developed during cycling at 55  C. thick. The 1st to 7th cycles are shown in this figure.

Toward the rechargeable capacity more than 300 mA h g
1 300 mA h g
1 could be observed even at room temperature when
the current is reduced to small value. Fig. 15 shows the discharge
The rechargeable capacity of Li[Li1/5Ni1/5Mn3/5]O2 remarkably
curves of Li[Li1/5Ni1/5Mn3/5]O2 examined at several currents at
increases by heating. Although there are some limiting factors in
room temperature. The cell was charged at 0.15 mA cm
2 up to
experiments, such as the stability of electrolyte, side reactions
5 V and then held at 5 V until the current decreased to 0.025 mA
during charge and discharge, etc., we would like to know the
cm
2, so-called CCCV mode. As clearly seen in Fig. 15, the
maximum rechargeable capacity observable for Li[Li1/5Ni1/5-
rechargeable capacity approaches 300 mA h g
1 when the current
Mn3/5]O2. The maximum rechargeable capacity observed in our
reduces from 1.65 to 0.075 mA cm
2 or 106.5 to 4.8 mA g
1. The
series of trials is 350 mA h g
1. The result is shown in Fig. 14. The
electrochemistry of Li[Li1/5Ni1/5Mn3/5]O2 suffers kinetic prob-
irreversible capacity observed at 85  C is 40 mA h g
1. The
lems which appear as polarization, hysteresis loop in charge and
steady-state charge and discharge curves are hardly observed
discharge curves, and variation of rechargeable capacity
when the cell is operated at 85  C. However, it is clear that Li[Li1/5-
depending on current drain together with temperature.
Ni1/5Mn3/5]O2 potentially has the rechargeable capacity of ca.
350 mA h g
1.
As stated above, the rechargeable capacity increases up to
350 mA h g
1 by heating. Such an effect of temperature upon the
rechargeable capacity strongly suggests that the electrochemical
reactions determining rechargeable capacity are under a kinetic
control. If this is true, the rechargeable capacity more than

Fig. 15 The effect of current upon discharge capacity of Li[Li1/5Ni1/5-

Mn3/5]O2 examined in a lithium cell at room temperature (ca. 25  C). The
cell was charged to 5.0 V at a constant current of 0.15 mA cm
2 (9.7 mA
g
1) and then kept at 5.0 V until the current decreased to 0.025 mA cm
2,
so called CCCV-charging mode, and then discharged at (a) 1.65 mA cm
2
Fig. 13 Differential chronopotentiograms calculated from numerical (106.5 mA g
1), (b) 1.05 (67.8), (c) 0.3 (19.4), (d) 0.15 (9.7), or (e) 0.075
data observed at (a) 2nd, (b) 10th, and (c) 20th cycle in Fig. 11. (4.8). Electrode mix used is 35.2 mg weight and 101 mm thick.

This journal is ª The Royal Society of Chemistry 2011 J. Mater. Chem., 2011, 21, 10179–10188 | 10185
 View Article Online

The reaction mechanism of Li[Li1/5Ni1/5Mn3/5]O2 showing the 1/2,Mn4+O2$1/2,Ni4+1/2Mn4+1/2O2 + Li+ + e
/ 1/2LiMn3+O2$

rechargeable capacity more than 300 mA h g
1 1/2LiNi2+1/2Mn4+1/2O2. (4)
Li[Li1/5Ni1/5Mn3/5]O2 shows the rechargeable capacity more The rechargeable capacity calculated from eqn (4) is 252 mA h
than 300 mA h g
1 and stably cycles with an average coulombic g
1 based on the weight of starting material of 1/2Li2Mn4+O3$1/2-
efficiency of 99.75% even at 55  C except the 1st cycle. The LiNi2+1/2Mn4+1/2O2. Therefore, the first charging capacity is
efficiency observed at the 1st cycle is 90%, which is usually expected to be 378 mA h g
1 and the rechargeable capacity is
observed for lithium insertion materials based on LiNiO2. The 252 mA h g
1. In other words, the irreversible capacity is once
charge and discharge curves at the 1st cycle are quite different observed to be 126 mA h g
1 and then the rechargeable capacity of
from those observed for lithium insertion materials. 252 mA h g
1 is observed for the subsequent cycles. Our results,
High-capacity positive-electrode materials based on lithium however, hardly support such a mechanism. So-called irreversible
Published on 11 April 2011. Downloaded by University of Leeds on 09/06/2013 06:55:23.

manganese oxides have already been reported and discussed the capacity observed is 70 mA h g
1 at room temperature, as seen in
possible mechanisms as to a long voltage plateau at 4.5 V and Fig. 8 and 9, or 40 mA h g
1 at 55  C in Fig. 10, which are too small
large specific capacity in mA h g
1.16–32 To describe the compared to the expected irreversible capacity of 126 mA h g
1.
composition under consideration, there seem to be two nota- The irreversible capacity observed at 55  C is smaller than that at
tions. One is Thackeray’s composite description in two- room temperature. In addition to the irreversible capacity, the
component notation of xLi2MnO3$(1
x)LiNi1/2Mn1/2O2,22 rechargeable capacity observed in this study is 230 mA h g
1 at
where 0 # x # 1, and the other is Dahn’s layered notation of Li room temperature, which may be explained because it is below the
[NiyLi(1/3
2y/3)Mn(2/3
y/3)]O2,17 where 0 # y # 1/2. In these expected rechargeable capacity of 252 mA h g
1. However, the
notations, the end members are LiNi1/2Mn1/2O213 and Li2MnO3 rechargeable capacities observed at 55 and 85  C are respectively
(Li[Li1/3Mn2/3]O2). Li[Li1/5Ni1/5Mn3/5]O2 used in this study is 300 and 350 mA h g
1, which are larger than the expected value of
equivalent to Thackeray’s notation with x ¼ 1/2 and Dahn’s 252 mA h g
1. The irreversible capacity smaller than 100 mA h g
1
notation with y ¼ 1/5. and the rechargeable capacity larger than 250 mA h g
1 cannot be
Among literatures there seem to be general agreements that explained using eqn (3) and (4).
(1) Li[Li1/5Ni1/5Mn3/5]O2 consists of divalent nickel and tetra- Oxygen evolution in eqn (3) supposedly occurs at the solid/
valent manganese ions in a cubic-close packed oxygen array electrolyte interface and consequently the surface composition
forming a layered structure, (2) a long voltage plateau at 4.5 V is changes reorganizing structure.31,32 The change in composition
observed in the first charging process, which is irreversible, and and structure at the solid/electrolyte interface reflects upon the
(3) the material shows considerably larger rechargeable capacity electrode potential. Transition metal ions may diffuse from the
compared to LiCoO2. We agree with these observations as have surface to bulk, changing the surface composition and conse-
been described in previous sections. quently the electrode potential. The kinetic process is sensitive to
The long voltage plateau at 4.5 V is usually explained in terms the previous electrode history. According to our results on
of irreversible loss of oxygen17 together with lithium removal window-opening experiments in Fig. 9 and 10, a 4.5 V plateau is
evolving O233,34 and reorganizing structure. According to a insensitive to previous history. When charging voltage reaches
proposed mechanism,31 the electrochemical extraction of lithium 4.5 V, the voltage always stays at that value until a reaction at 4.5
from Li[Li1/5Ni1/5Mn3/5]O2 or alternatively 1/2Li2MnO3$1/2- V is completed. Then, question arises as to the irreversible loss of
LiNi1/2Mn1/2O2 occurs in two steps, i.e., oxygen from Li[Li1/5Ni1/5Mn3/5]O2.
The irreversible loss of oxygen is first proposed by Lu and
1/2Li2Mn4+O3$1/2LiNi2+1/2Mn4+1/2O2 / 1/2Li2Mn4+O3$1/2, Dahn17 in their electrochemical and structural study of Li
Ni4+1/2Mn4+1/2O2 + 1/2Li+ + 1/2e
(2) [NixLi(1/3
2x/3)Mn(2/3
x/3)]O2 for x ¼ 1/6, 1/4, 1/3, 5/12, and 1/2.
To find one-to-one correspondence between electrochemical
and and structural data before and after a long voltage plateau,
they use electrochemical data to calculate the composition of
1/2Li2Mn4+O3$1/2,Ni4+1/2Mn4+1/2O2 / 1/2,Mn4+O2$1/2, lithium ions in their samples, and then its composition is fixed
Ni4+1/2Mn4+1/2O2 + Li+ + 1/4O2 + e
. (3) in analyzing in situ XRD results while oxygen positional
parameter and the occupation of the oxygen ions are set to be
Reaction (2) is reversible in voltage below 4.5 V and reaction (3) variable. They find that the best fit is obtained when the
is irreversible at the 4.5 V plateau. The theoretical capacity occupation of the oxygen sites is below 100% for the electrodes
calculated from eqn (2) based on the sample weight is 126 mA h charged at 4.8 V, resulting in the description on the loss of
g
1, which well agrees with experimental results in Fig. 9 and 10. oxygen. Our ex situ XRD results examined using Fe Ka-
The capacity due to reaction (3) is calculated to be 252 mA h g
1 to radiation are consistent with their observations. After the first
form new active material of 1/2,Mn4+O2$1/2,Ni4+1/2Mn4+1/2O2 charge to 5 V, the lattice volume contracts below the original
in the long plateau at 4.5 V. The observed values range from 180 lattice volume. During subsequent charge and discharge, the
mA h g
1 at room temperature in Fig. 8 to 270 mA h g
1 at 85  C in lattice volume expands on discharge and contracts on charge.
Fig. 14. These values may be explained by the above argument After the first cycle, the lattice volume is irreversibly expanded
based on eqn (3) if possible side reactions above 4.5 V at high to larger value than that of the original one. The a-axis shrinks
temperature are taken into account. on charge and elongates on discharge. The c-axis elongates on
After first charging to 5 V, 1/2,Mn4+O2$1/2,Ni4+1/2Mn4+1/2- charge until it reaches a certain value and then shrinks a little.
O2 is presumably new active material for subsequent cycles, i.e., XRD observations are quite similar to those of

10186 | J. Mater. Chem., 2011, 21, 10179–10188 This journal is ª The Royal Society of Chemistry 2011
 View Article Online

LiCoxNi1
xO2.35,36 As clearly seen in Fig. 4 and 6 combined Another possibility is an anion redox reaction in a solid
with in situ XRD observation in ref. 17, the electrochemical matrix. If we assume the redox reaction of 2O2
/O22
substituting
reactions of a series of Li[NixLi(1/3
2x/3)Mn(2/3
x/3)]O2 seem- for eqn (6), the reaction is represented by
ingly proceed in a topotactic manner. No dramatic change in
XRD and neutron diffraction33 can be seen and the voltage ,2/5Li3/5[Li1/5Ni4+1/5Mn4+3/5]O2 / ,[Li1/5Ni4+1/5Mn4+3/5]
during charge and discharge monotonously changes as a func- (2O )7/10(O22
)3/10 + 3/5e
+ 3/5Li+,
2

(8)
tion of the degree of reaction, so that the reactions are clas-
sified as a homogeneous phase or one-phase reaction under the and it is further oxidized by
framework of lithium insertion electrodes.
Flat operating voltage on the first charge reminds us the case for ,[Li1/5Ni4+1/5Mn4+3/5](2O2
)7/10(O22
)3/10 / ,[,1/5Ni4+1/5-
LiVO237 or LT-LiMnO2.38–40 The original crystal lattices are Mn4+3/5](2O2
)3/5(O22
)2/5 + 1/5e
+ 1/5Li+. (9)
Published on 11 April 2011. Downloaded by University of Leeds on 09/06/2013 06:55:23.

entirely destroyed at flat operation voltage, which is the conven-

The theoretical capacities based on eqn (8) and (9) are the same
tional two-phase reaction. LiVO2 is converted to an inactive
as previously calculated using eqn (6) and (7) because the number
layered material and LiMnO2 is converted to an ill-defined spinel
of electrons transferred per a formula unit of Li[Li1/5Ni1/5Mn3/5]
phase, which is active for subsequent cycles with small irreversible
O2 is the same. Peroxide ions in a solid matrix are seen in Na2O2,
capacity. The hysteresis loops observed during charge and
which is a reaction product of sodium oxidation with dry oxygen,
discharge are similar to that of ill-defined spinel phase derived
and also in ZnO2 consisting of Zn2+ and O22
. They are stable in
from LiMnO2.38–41 The hysteresis loop may be explained in terms
non-aqueous environment. An anion redox reaction in a solid
of a kinetic effect, such as slow lithium diffusion and migration in
matrix is not known for oxygen but known for sulfur. Pyrite FeS2
a solid matrix. The slow process is usually accelerated when an
consisting of Fe2+ and S22
is reduced to Li2FeS2 consisting of
electrode system is heated. However, the hysteresis loop observed
Li+, Fe2+ and 2S2
. The reaction is reversible. The theoretical
at 55  C in Fig. 10 is larger than that at room temperature in Fig. 8.
capacity is calculated to be 447 mA h g
1, which is two times
We need to find a possible explanation for a series of mysterious
larger than that calculated for LiMeS2 4 ,MeS2 + Li+ + e

phenomena observed for Li[NixLi(1/3
2x/3)Mn(2/3
x/3)]O2, espe-
because of two-electron redox reaction of anions, i.e., S22
+ 2e

cially Li[Li1/5Ni1/5Mn3/5]O2.
/ 2S2
.43
If we assume penta-valent manganese ions surrounded by six or
The above arguments are based on ‘‘cation’’ or ‘‘anion’’ redox
four oxygen ions, a solid state redox reaction of Mn4+/Mn5+,
reaction in a solid matrix. Such an elementary step of the reaction
specifically MnO68
/MnO67
or MnO68
/MnO43
in a local
based on the oxidation number does not hold in a solid-state
environment, may participate in the reaction under considera-
redox reaction, as was discussed previously.5 During an oxida-
tion, i.e.,
tion of LiMeO2, electrons are removed from oxygen ions rear-
ranging electron distribution between anions and cations, and
Li[Li1/5Ni2+1/5Mn4+3/5]O2 / ,2/5Li3/5[Li1/5Ni4+1/5Mn4+3/5]O2 +
consequently transition metal ions are sometimes reduced in
2/5e
+ 2/5Li+, (5)
spite of the oxidation of LiMeO2.
which is another expression of reversible reaction (2), and then
Conclusion
,2/5Li3/5[Li1/5Ni 4+
1/5Mn
4+
3/5]O2 / ,[Li1/5Ni 4+
1/5Mn
5+
3/5]O2+ In this paper, we have dealt with high-capacity lithium nickel
3/5e
+ 3/5Li+. (6) manganese dioxides and shown that the layered material of Li
[Li1/5Ni1/5Mn3/5]O2 potentially has the rechargeable capacity of
The theoretical capacities based on eqn (5) and (6) are 350 mA h g
1, which clearly exceeds the critical limit of 300 mA h
respectively 126 and 189 mA h g
1. Totally 315 mA h g
1 is g
1 with respect to the previous concept of lithium insertion
expected for the one-electron transfer based on the original materials based on a topotactic reaction of LiMeO2 4 Li+ + e
+
sample of Li[Li1/5Ni2+1/5Mn4+3/5]O2. Further oxidation is possible ,MeO2. Although the irreversible reaction at a plateau voltage
if we assume the formation of Mn6+ surrounded by four oxygen of 4.5 V observed at first charge is still a debatable subject, all
ions, MnO42
, because lithium ions to be removed are experimental results can be explained in terms of lithium inser-
available in eqn (6). Then, an addition capacity is calculated to be tion scheme. Understanding of solid-state chemistry and elec-
63 mA h g
1 by trochemistry underlying the rechargeable capacity more than
300 mA h g
1 will open a new window toward an innovative
,[Li1/5Ni4+1/5Mn5+3/5]O2 / ,[,1/5Ni4+1/5Mn5+2/5Mn6+1/5]O2 + concept on lithium insertion materials for advanced lithium-ion
1/5e
+ 1/5Li+. (7) batteries. We believe that this paper will help break through the
conventional limit of concept on lithium insertion materials.
If we account for eqn (7), totally 378 mA h g
1 is expected.
Chemical species of manganese ions having higher oxidation
number than 4+ are known as green manganate ions of
Acknowledgements
MnO42
and purple permanganate ions of MnO4
. Blue The authors wish to thank previous graduate students,
manganese ions may be the oxidation number of 5+.42 These Mr Takaaki Izumoto and Mr Kenichi Honoki, for their help on
species are observed in strong acid or base under the preparing and examining some of the samples. The present work
oxidative condition, such as concentrated H2SO4 or KOH was partially supported by a grant-in-aid for scientific research
under O2. from the Osaka City University Science Foundation.

This journal is ª The Royal Society of Chemistry 2011 J. Mater. Chem., 2011, 21, 10179–10188 | 10187
 View Article Online

References 24 P. S. Whitfield, S. Niketic and I. J. Davidson, J. Power Sources, 2005,

146, 617.
1 T. Ohzuku and A. Ueda, Solid State Ionics, 1994, 69, 201. 25 J. Breger, M. Jiang, M. Dupre, Y. S. Meng, Y. Shao-Horn, G. Ceder
2 M. Winter, J. O. Besenhard, M. E. Spahr and P. Novak, Adv. Mater., and C. P. Grey, J. Solid State Chem., 2005, 178, 2575.
1998, 10, 725. 26 C. S. Johnson, N. Li, J. T. Vaughey, A. A. Hackney and
3 B. Ammundsen and J. Paulsen, Adv. Mater., 2001, 13, 943. M. M. Thackeray, Electrochem. Commun., 2005, 7, 528.
4 M. S. Whittingham, Chem. Rev., 2004, 104, 4271. 27 D.-K. Lee, S.-H. Park, K. Amine, H. J. Bang, J. Parakash and
5 T. Ohzuku, K. Ariyoshi, Y. Makimura, N. Yabuuchi and K. Sawai, Y.-K. Sun, J. Power Sources, 2006, 162, 1346.
Electrochemistry (Tokyo, Jpn.), 2005, 73, 2. 28 C. S. Johnson, N. Li, C. Lefief and M. M. Thackeray, Electrochem.
6 T. Ohzuku and R. Brodd, J. Power Sources, 2007, 174, 449.
Commun., 2007, 9, 787.
7 B. L. Ellis, K. T. Lee and L. F. Nazar, Chem. Mater., 2010, 22, 691.
29 T. A. Arunkumar, E. Alvarez and A. Manthiram, J. Electrochem.
8 T. Ohzuku, A. Ueda, M. Nagayama, Y. Iwakoshi and H. Komori,
Electrochim. Acta, 1993, 38, 1159. Soc., 2007, 154, A770.
9 T. Ohzuku, A. Ueda and M. Kouguchi, J. Electrochem. Soc., 1995, 30 A. Ito, D. Li, Y. Ohsawa and Y. Sato, J. Power Sources, 2008, 183,
Published on 11 April 2011. Downloaded by University of Leeds on 09/06/2013 06:55:23.

142, 4033. 344.

10 K. Ariyoshi, Y. Iwakoshi, N. Nakayama and T. Ohzuku, 31 S.-H. Kang and M. M. Thackeray, Electrochem. Commun., 2009, 11,
J. Electrochem. Soc., 2004, 151, A296. 748.
11 P. Poizot, S. Laruelle, S. Grugeon, L. Dupont and J.-M. Tarascon, 32 A. Ito, D. Li, Y. Sato, M. Arao, M. Watanabe, M. Hatano, H. Horie
Nature, 2000, 407, 496. and Y. Ohsawa, J. Power Sources, 2010, 195, 567.
12 A. S. Arico, P. Bruce, B. Scrosati, J.-M. Tarascon and W. Van 33 A. R. Armstrong, M. Holzapfel, P. Novak, C. S. Johnson,
Schalkwijk, Nat. Mater., 2005, 4, 366. S.-H. Kang, M. M. Thackeray and P. G. Bruce, J. Am. Chem. Soc.,
13 H. Ito, M. Shimakawa, T. Tanaka and T. Ohzuku, Extended Abstract 2006, 128, 8694.
on IBA-HBC2003 Ernest Yeager Memorial Symposium, Hawaii, Jan 34 M. Holzapfel, A. W€ ursig, W. Sheifele, J. Vetter and P. Novak,
2003, p. 83. J. Power Sources, 2007, 174, 1156.
14 C. Kittel, in Introduction to Solid State Physics, John Wiley & Sons, 35 A. Ueda and T. Ohzuku, J. Electrochem. Soc., 1994, 141, 2010.
Inc., New York, 6th edn, 1986, p. 531. 36 T. Ohzuku and Y. Makimura, Res. Chem. Intermed., 2006, 32, 507.
15 T. Ohzuku and Y. Makimura, Chem. Lett., 2001, 30, 744. 37 L. A. de Picciotto, M. M. Thackeray, W. I. F. David, P. G. Bruce and
16 C. S. Johnson, J.-S. Kim, A. J. Kropf, A. J. Kahaian, J. T. Vaughey J. B. Goodenough, Mater. Res. Bull., 1984, 19, 1497.
and M. M. Thackeray, Electrochem. Commun., 2002, 4, 492. 38 T. Ohzuku, A. Ueda and T. Hirai, Chem. Express, 1992, 7, 193.
17 Z. Lu and J. R. Dahn, J. Electrochem. Soc., 2002, 149, A815. 39 J. N. Remers, E. W. Fuller, E. Rossen and J. R. Dahn, J. Electrochem.
18 Z. Lu and J. R. Dahn, J. Electrochem. Soc., 2002, 149, A1454.
Soc., 1993, 140, 3396.
19 C. S. Johnson, J.-S. Kim, A. J. Kropf, A. J. Kahaian, J. T. Vaughey
40 I. M. K€otschau and J. R. Dahn, J. Electrochem. Soc., 1998, 145, 2672.
and M. M. Thackeray, J. Power Sources, 2003, 119–121, 139.
41 T. Ohzuku, S. Kitano, M. Iwanaga, H. Matsuno and A. Ueda,
20 G. J. Moore, C. S. Johnson and M. M. Thackeray, Power Sources,
2003, 119–121, 216. J. Power Sources, 1997, 68, 646.
21 Z. Lu, Z. Chen and J. R. Dahn, Chem. Mater., 2003, 15, 3214. 42 M. Pourbaix, Atlas of Electrochemical Equilibria in Aqueous Solutions,
22 C. S. Johnson, J.-S. Kim, C. Lefief, N. Li, J. T. Vaughey and Pergamon Press, New York, 1966.
M. M. Thackeray, Electrochem. Commun., 2004, 6, 1085. 43 J. Barker and E. Kendrick, Abstract #435, The 15th International
23 Y. S. Meng, G. Ceder, C. P. Grey, W.-S. Yoon, M. Jiang, J. Breger Meeting on Lithium Batteries, Montreal, Canada, June 27 to July 2,
and Y. Shao-Horn, Chem. Mater., 2005, 17, 2386. 2010.

10188 | J. Mater. Chem., 2011, 21, 10179–10188 This journal is ª The Royal Society of Chemistry 2011