Nanostructured lithium nickel manganese oxides for lithium-ion batteries

Article abstract of DOI:10.1149/1.3308598

Nanostructured lithium nickel manganese oxides were investigated as advanced positive electrode materials for lithium-ion batteries designated to power plug-in hybrid electric vehicles and all-electric vehicles. The investigation included material characterization and electrochemical testing. In cell tests, the Li_1.375 Ni_0.25 Mn_0.75 O _2.4375 composition achieved high capacity (210 mAh g^-1) at an elevated rate (230 mA g^-1), which makes this material a promising candidate for high energy density Li-ion batteries, as does its being cobalt-free and uncoated. The material has spherical morphology with nanoprimary particles embedded in micrometer-sized secondary particles, possesses a multiphase character (spinel and layered), and exhibits a high packing density (over 2 g cm^-3) that is essential for the design of high energy density positive electrodes. When combined with the Li₄ Ti ₅ O₁₂ stable anode, the cell showed a capacity of 225 mAh g^-1 at the C/3 rate (73 mA g^-1) with no capacity fading for 200 cycles. Other chemical compositions, Li_(1+x) Ni _0.25 Mn_0.75 O_(2.25+x/2) (0.32 ≤ x ≤ 0.65), were also studied, and the relationships among their structural, morphological, and electrochemical properties are reported.

Full text of DOI:10.1149/1.3308598

Journal of The Electrochemical Society, 157 ͑4͒ A447-A452 ͑2010͒
A447
0
013-4651/2010/157͑4͒/A447/6/$28.00 © The Electrochemical Society
Nanostructured Lithium Nickel Manganese Oxides
for Lithium-Ion Batteries
a,
a, ,z
b
a,
Haixia Deng, * Ilias Belharouak, * Russel E. Cook, Huiming Wu, *
c,
a, ,z
Yang-Kook Sun, * and Khalil Amine *
a
b
Chemical Sciences and Engineering Division and Materials Science Division, Argonne National
Laboratory, Argonne, Illinois 60439, USA
c
Center for Information and Communication Materials, Department of Chemical Engineering, Hanyang
University, Seoul 133-791, Korea
Nanostructured lithium nickel manganese oxides were investigated as advanced positive electrode materials for lithium-ion
batteries designated to power plug-in hybrid electric vehicles and all-electric vehicles. The investigation included material char-
acterization and electrochemical testing. In cell tests, the Li1.375Ni0.25Mn0.75O2.4375 composition achieved high capacity
͑
210 mAh g−1͒ at an elevated rate ͑230 mA g ͒, which makes this material a promising candidate for high energy density
−1
Li-ion batteries, as does its being cobalt-free and uncoated. The material has spherical morphology with nanoprimary particles
embedded in micrometer-sized secondary particles, possesses a multiphase character ͑spinel and layered͒, and exhibits a high
packing density ͑over 2 g cm−3͒ that is essential for the design of high energy density positive electrodes. When combined with
−
1
−1
the Li Ti O stable anode, the cell showed a capacity of 225 mAh g at the C/3 rate ͑73 mA g ͒ with no capacity fading for
4
5
12
2
00 cycles. Other chemical compositions, Li͑1+x͒Ni0.25Mn0.75O͑2.25+x/2͒ ͑0.32 Յ x Յ 0.65͒, were also studied, and the relationships
among their structural, morphological, and electrochemical properties are reported.
2010 The Electrochemical Society. ͓DOI: 10.1149/1.3308598͔ All rights reserved.
©
Manuscript submitted November 24, 2009; revised manuscript received December 22, 2009. Published February 25, 2010.
Rechargeable lithium-ion batteries were first commercialized by
Sony in the early 1990s. The high cost of cobalt and the relatively
and surface modification as routes to improve the electronic conduc-
15-17
tivity of the material and thus enhance its rate capability.
−
1
low specific capacity of LiCoO ͑ϳ140 mAh g ͒ have been major
Our approach is based on enhancing the power capability of
Li1.5Ni0.25Mn0.75O2.5 without jeopardizing its volumetric energy
density. In general, the electrochemical performance of an electro-
active material strongly depends not only upon its intrinsic proper-
ties such as the chemical composition and crystalline structure, but
also upon its textural and morphological properties, which are
greatly dependent on the fabrication methods available in the field of
materials science. Drawing upon state-of-the-art methods, we syn-
2
obstacles against the application of these batteries in plug-in hybrid
vehicles ͑PHEVs͒ and all-electric vehicles ͑EVs͒. Also, despite the
abundance of iron and manganese, olivine LiFePO4 and spinel
LiMn O are not likely adequate candidates because neither of them
2
4
provides high enough gravimetric and volumetric energy densities
for transportation applications.
Therefore, research groups have been inventing and developing
cathode materials whose main characteristic is to store more and
more electricity per mass and volume. In 2001, Ohzuku and
thesized nanostructured Li͑1+x͒Ni0.25Mn0.75
O
materials
͑2.25+x/2͒
͑0.32 Յ x Յ 0.65͒ using a Ni0.25Mn0.75CO3 carbonate precursor
prepared via a coprecipitation method in a continuous stirred tank
reactor. By tuning the lithium concentration, it was possible to moni-
tor the sizes of the nanodomains that are embedded in the large
spherical micrometer-sized mesoporous particles. Primarily because
of their physical properties, these materials provide a higher electro-
active surface for faster lithium-ion diffusion and a higher packing
density for greater energy storage. The structural, physical, and elec-
trochemical properties of these materials are reported and discussed
below.
1,2
Makimura and Lu et al. reported cathode materials with more than
−
1
2
00 mAh g specific capacity. These materials attracted signifi-
cant attention because they also exhibit much improved thermal
2
stability. Of these materials, Li1.5Ni0.25Mn0.75O2.5 ͑for convenience,
also written as Li1.2Ni0.2Mn0.6O ͒ delivers a high reversible capacity
2
3
,4
while being structurally stable upon charging to 4.6 V. These
properties, in addition to its being cobalt-free, make
Li1.5Ni0.25Mn0.75O2.5 an excellent candidate to surmount the energy
density shortfall of present lithium-ion batteries and to meet the
lower cost and low toxicity requirement targeted by industry.
Several analysis tools and models have been used to better un-
Experimental
derstand
the
electrochemical
mechanism
by
which
A two-step synthesis process was adopted to make nanosized
Li1.5Ni0.25Mn0.75O2.5 can deliver a capacity exceeding the electro-
chemical activity of Ni ions.
proaches has focused on the Li MnO component and its role in the
deliverance of additional capacity through the removal of Li O.
In parallel to this fundamental research, there has been considerable
interest in demonstrating the practical use of this material because
the low volumetric energy density and poor rate capability are still
not resolved. Recourse to nanoscale materials has been viewed as a
means to take full advantage of the specific capacity of the material
at a high rate but with a drawback of low packing densities
structured active materials Li͑1+x͒Ni0.25Mn0.75
O
͑where x
͑2.25+x/2͒
2
+
5-11
One of the more promising ap-
=
0.32, 0.375, 0.5, 0.575, and 0.65͒. The first step consisted of syn-
2
3
thesizing a spherical Ni0.25Mn0.75CO carbonate precursor through a
9
,11
3
2
coprecipitation process. The preparation process of the precursor
18
was detailed in our previous work. The resulting Ni0.25Mn0.75CO3
powder was dried inside a vacuum oven at 100°C for several hours.
Thereafter, stoichiometric amounts of Li CO ͑Ն99.0%, Aldrich͒
2
3
were intimately mixed with the as-prepared Ni0.25Mn0.75CO car-
3
bonate particles. Calcination was accomplished by a first treatment
at 600°C for 15 h under air to decompose the carbonate precursors,
followed by 900°C for 15 h to obtain the corresponding lithiated
compounds.
The prepared samples were analyzed by various analytical meth-
ods. A powder X-ray diffractometer ͑XRD, Siemens D5000͒ was
used to analyze the crystal structure and the phase composition for
both the carbonate precursor and the lithiated compounds. The ra-
diation source was Cu K␣. The samples were scanned from 2␪
−
3
͑
0.5–0.8 g cm ͒ and, as
a
result, low volumetric energy
1
2,13
−
densities.
Moreover, Li1.5Ni0.25Mn0.75O2.5 provides over
1
200 mAh g capacity and reasonable cycling performances only
−
1
under low rate conditions, e.g., C/25 ͑10 mA g ͒ or C/50
−
1
4,5,11,14
͑
5 mA g ͒.
Other researchers have pursued cobalt doping
=
10 to 100° at a low scan rate of 20 s per 0.02°.
Scanning electron microscopy ͑SEM͒, transmission electron mi-
*
Electrochemical Society Active Member.
E-mail: belharouak@anl.gov; amine@anl.gov
z
Downloaded on 2014-12-21 to IP 130.209.6.50 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A448
Journal of The Electrochemical Society, 157 ͑4͒ A447-A452 ͑2010͒
SEM image shows a solid texture inside the particles ͓inset ͑c͔͒ with
a Ni:Mn transition-metal ratio close to the 1:3 nominal ratio ͑EDXS
result not shown here͒. These morphological characteristics led to a
(a)
−
3
10 µm
carbonate precursor having a high packing density of 1.8 g cm
with high dispersibility and spherical properties that are indispens-
able for better mixing during the addition of lithium carbonate and
hence for enhanced reactivity and homogeneity during calcination.
(
b)
The Ni0.25Mn0.75CO precursor was dried, and a thermal gravimetric
analysis was conducted to ensure the material’s complete dryness
3
1
µm
(
c)
before mixing with dry Li CO . In early experiments, reproducibil-
2
3
ity issues were encountered because of slight variations in lithium
contents in undried materials. Since then, we intentionally intro-
duced an increasing amount of dry Li CO to dry Ni0.25Mn0.75CO3,
10 µm
2
3
according to the formula, Li͑1+x͒Ni0.25Mn0.75
O
͑x = 0.32,
͑2.25+x/2͒
0
.375, 0.5, 0.575, and 0.65͒. The oxygen stoichiometry, as written
for these compounds, was calculated to balance the total positive
charge
+
1
+
2
0
30
40
50
60
70
80
in
Li͑1+x͒Ni Mn0.75
0.25
O
͑2.25+x/2͒
͑͑1 + x͒·1͓Li ͔
2
+
4+
0.25·2͓Ni ͔ + 0.75·4͓Mn ͔͒ according to the most stable oxida-
2
θ
tion states of lithium, nickel, and manganese in similar cathode
2
-12
materials.
trometry was used to track the chemical composition in all samples.
Figure shows the XRD patterns of the lithiated
Li͑1+x͒Ni0.25Mn0.75 compounds: ͑a͒ Li1.32Ni0.25Mn0.75O2.41
Inductively coupled plasma/optical emission spec-
Figure 1. XRD pattern of the Ni0.25Mn0.75CO3 carbonate precursor. The
insets ͑a͒, ͑b͒, and ͑c͒ are the SEM images of secondary particles, primary
2
particles, and cross section of Ni0.25Mn0.75CO , respectively.
3
O
,
͑2.25+x/2͒
͑
b͒ Li1.375Ni0.25Mn0.75O2.4375
,
͑c͒ Li1.5Ni0.25Mn0.75O2.5
,
͑d͒
Li1.575Ni0.25Mn0.75O2.5375, and ͑e͒ Li1.65Ni0.25Mn0.75O2.575. These
notations have been somewhat simplified in the literature to reflect a
croscopy ͑TEM͒, and focused ion beam ͑FIB͒ specimen prepara-
tions were done in the Electron Microscopy Center for Materials
Research, Argonne National Laboratory. The SEM system used for
the research was a Hitachi S-4700-II with a cold field emitter, a
GW-type backscattered electron ͑BSE͒ detector, and an energy-
dispersive X-ray spectrometer ͑EDXS͒. The TEM used for imaging
and selected area electron diffraction ͑SAD͒ consisted of an FEI
specific structural order. For instance,
a material such as
Li1.5Ni0.25Mn0.75O2.5 ͓spectrum ͑c͔͒ could be written as
Li1.2Ni0.2Mn0.6O2 by dividing by a factor of 1.25 to reflect the no-
tation Li͑Li0.2Ni0.2Mn0.6͒O2 that is consistent with the structural or-
¯
der of the layered ␣-NaFeO ͑R3m͒. An advanced notation has been
2
CM30T with a LaB emitter, an EDXS, and a Gatan 666 parallel
6
widely adopted to reflect the structural complexity and the compos-
ite character of these materials. For example, Li1.5Ni0.25Mn0.75O2.5
could be written as ͓1/2Li MnO ·1/2LiNi0.5Mn0.5O ͔ to show that
electron energy-loss spectrometer. The SEM and TEM specimens
were prepared in Zeiss cross-beam FIB instruments ͑NVision and
2
3
2
1
540XB͒. The packing density was measured by a standard method
¯
using an Autotap instrument from Quantachrome Instruments. The
packing density is a measurement of how much weight of a given
material can be packed per volume without applying any kinds of
pressure other than taping until no further volume reduction can be
observed.
two layered components ͑C2/m and R3m͒ can be structurally inte-
7,9
grated to form a composite compound. Although we agree with
the composite formulation, we decided to write the chemical com-
position of our materials as Li͑1+x͒Ni0.25Mn0.75O͑2.25+x/2͒ without
preference to any structural order notations. In the rest of the paper,
the studied samples are labeled as ͑a͒, ͑b͒, ͑c͒, ͑d͒, and ͑e͒ corre-
sponding to x values of 0.32, 0.375, 0.5, 0.575, and 0.65, respec-
tively.
Electrochemical measurements were carried out on CR2032-type
2
coin cells ͑1.6 cm ͒. The positive electrodes were made of 80 wt %
active oxide materials, 10 wt % acetylene black as the conductive
agent, and 10 wt % poly͑vinylidene difluoride͒ binder. The loading
The XRD patterns of these samples do not fit the description of a
one-layer structural model, namely, the structural model of
2
density of the active material was around 5 mg/cm . The electrolyte
was 1.2 M LiPF dissolved in a mixture of ethylene carbonate ͑EC͒
¯
6
␣-NaFeO ͑R3m͒ or that of Li MnO ͑C2/m͒. They are composed
2
2
3
and ethyl methyl carbonate ͑EMC͒ in a 3:7 volume ratio. The cells
were assembled with lithium metal as the negative electrode and
were tested in the voltage ranges of 2–4.6 and 2–4.9 V.
Differential scanning calorimetry ͑DSC͒ using a Perkin-Elmer
Pyris-1 instrument was conducted on electrochemically delithiated
electrodes ͑slowly charged to 4.6 V at the C/40 rate and held at 4.6
V for another 40 h for equilibrium͒. Around 3 mg of scraped elec-
trode materials and 3 ␮L of electrolyte were hermetically sealed
inside stainless steel high pressure capsules. The DSC curves were
recorded between room temperature and 375°C at a scan rate of
of at least these components in addition to a third one that was
observed with materials having less lithium content ͓Fig. 2, samples
͑
a͒ and ͑b͔͒. Indeed, a close examination of the pattern of sample ͑a͒
shows the presence of an unusual broadening on the left side of the
peaks observed at 36.92 and 44.58°. This broadening is less pro-
nounced in sample ͑b͒ and visually disappears for samples ͑c͒, ͑d͒,
and ͑e͒. Figure 3 shows the details of the XRD pattern of sample ͑a͒
where the broadening is evident. However, it was not possible for us
¯
to assign this broadening to R3m or C2/m space groups with great
confidence. Within a slight peak position variation, the X-ray lines
10°C/min.
¯
arising from space group R3m could easily be hidden/overlapped by
Results and Discussion
Crystal structure.— Figure 1 shows the XRD pattern of the
the lines arising from space group C2/m, and the same could be said
¯
¯
for Fd3m, for which all lines could be hidden/overlapped by R3m.
As a first approximation, as in other research, by ignoring the
possible contribution of the layered component ͑C2/m͒ and the spi-
Ni0.25Mn0.75CO carbonate precursor. The pattern fits well with the
3
structure of MnCO , which attests to the purity of the starting pre-
3
cursor. The insets in Fig. 1 show the SEM images of the particles by
¯
nel ͑Fd3m͒, we were able to use Rietveld profile matching refine-
ment to calculate the unit cell parameters of all samples based on the
which Ni0.25Mn0.75CO is constituted. Dense and spherical second-
3
ary particles can be observed with an average particle size of
¯
11,19
1
5 ␮m ͓inset ͑a͔͒. At high magnification, the image of the surface
most significant layered component ͑R3m͒ ͑Fig. 2, right side͒.
of the secondary particles revealed the presence of highly packed
Within this hexagonal structural model, diffraction parameters ͑a͒
and ͑c͒ had their values slightly increased from sample ͑a͒ to sample
nanosized primary agglomerates ͓inset ͑b͔͒. The cross-sectional
Downloaded on 2014-12-21 to IP 130.209.6.50 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, 157 ͑4͒ A447-A452 ͑2010͒
A449
2
.860
.855
.850
.845
Li MnO -like
2
3
2
peaks
broadening
broadening
(a)
2
2
2
.840
14.255
4.250
4.245
4.240
4.235
4.230
4.225
00.8
broadening
broadening
(b)
1
1
1
1
1
1
1
Figure 2. XRD patterns of
͑
a͒ Li1.32Ni0.25Mn0.75O2.41
,
͑b͒ Li1.375Ni0.25Mn0.75
O
,
2.4375
͑
͑
͑
c͒ Li1.5Ni0.25Mn0.75O2.5
,
d͒ Li1.575Ni0.25Mn0.75O2.5375
e͒ Li1.65Ni0.25Mn0.75O2.575 ͑left side͒, and
(c)
,
corresponding lattice parameter evolution
as function of lithium content ͑right side͒.
(
d)
1
00.6
00.4
00.2
00.0
1
(e)
1
1
1
0
20
30
40
50
60
70
80
1.30 1.35 1.40 1.45 1.50 1.55 1.60 1.65
ο
XinLi1+xNi0.25Mn0.75O
2.25+x/2
2
θ,
͑
c͒ and then slightly decreased from sample ͑c͒ to sample ͑e͒, with a
Physical characteristics.— Figure 4 shows SEM images of the
lithiated samples ͑a͒–͑e͒. The photographs on the left side show that
all compounds kept the spherical morphology and the size of the
similar trend of the cell volume for which the value increased and
decreased accordingly ͑Fig. 2, right side͒. As for the meaning of
these observations, we have not pursued the traditional approach of
developing a single explanation based on the ionic radii of
Li ͑0.76 Å͒, Ni ͑0.69 Å͒, and Mn ͑0.53 Å͒ along with
crystal considerations. Because there is evidence of other crystalline
Ni0.25Mn0.75CO precursor. However, as apparent on the right side
3
of Fig. 4, the primary particles that construct the secondary particles
underwent significant size change after lithiation, namely, from
around 80 nm ͓sample ͑a͔͒ to around 200–500 nm for samples with
the highest lithium content ͓͑d͒ and ͑e͔͒. The images also show that
the primary particles that compose the surface of samples ͑a͒ and ͑b͒
form a much denser and smoother surface than that of the other
samples, with much more pronounced roughness for the samples
containing higher lithium concentrations ͑Fig. 4d and e͒. These mor-
phological observations were accompanied by a gradual drop in the
+
2+
4+
¯
components, i.e., the spinel component ͑Fd3m͒ for the lower
lithium content ͑see electrochemical evidence below͒ and the lay-
ered component ͑C2/m͒ that exists for all lithium concentrations,
there are possible X-ray peak overlaps and slight repositioning that
might lead to false parameter calculations. In addition to the com-
posite characteristic of these materials at the atomic level, as sug-
gested and demonstrated by Thackeray et al., a simple physical mix
−
3
packing density from the highest value of 2.1 g cm ͓samples ͑a͒
and ͑b͔͒ to the lowest value of 1.7 g cm− ͓samples ͑d͒ and ͑e͔͒.
Samples ͑a͒ and ͑b͒ exhibited a relatively higher surface area ͑
3
¯
is possible at the particle level between the components Fd3m and
¯
9,20
2
2
R3m.
ϳ8.9 m /g͒ compared to samples ͑c͒, ͑d͒, and ͑e͒ ͑ϳ4 m /g͒. The
nanoparticles ͑or primary particles͒ that form the secondary particles
are the result of the calcination process in which the carbonate pre-
cursor Ni0.25Mn0.75CO decomposes to an oxide by releasing CO
3
2
Layered C2/m
Layered R-3m
Spinel Fd-3m
Sample (a)
and creating voids inside the particles. Depending on the lithium
amounts, these nanoparticulates can grow more or less during lithia-
tion. In general, the lower the lithium content, the smaller the pri-
mary particles and the higher the compactability.
spinel
To further advance this morphological study, a particle from
sample ͑a͒ was Ga-milled at an angle of 36° from the plane of the
aluminum sample stub on which it sat. The particle was tilted so that
the flat surface of the milled particle was normal to the direction of
the SEM electron beam before the acquisition of the BSE image.
Figure 5 is a BSE image obtained with an accelerating voltage of 15
kV. The particle has an obvious core and shell configuration, with a
layer separating the two major regions. The cracks are intrinsic to
the particle. Figure 6a is a diffraction-contrast bright-field image of
an FIB-prepared cross section from this particle. The core of the
particle is at the lower left, and the shell is at the upper right. The
separating layer bisects the image, curving from the upper left to the
lower right. The image shows that the particle is both nanocrystal-
line ͑by diffraction contrast͒ and nanoporous ͑white, featureless re-
gions͒. Figure 6b is an SAD pattern from the shell region. It shows
that the shell, while polycrystalline, has a preferred-growth direc-
tion. However, the core is polycrystalline and lacks a preferred-
spinel
34
36
38
40
42
44
46
Fd-3m
R-3m
C2/m
15
20
25
30
35
40
45
50
55
60
2
θ
Figure 3. ͑Color online͒ XRD pattern of sample ͑a͒ Li1.32Ni0.25Mn0.75O2.41
highlighting the existence of a third phase.
,
Downloaded on 2014-12-21 to IP 130.209.6.50 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A450
Journal of The Electrochemical Society, 157 ͑4͒ A447-A452 ͑2010͒
(a)
(a’)
Shell
Core
(
b)
(b’)
(c’)
(d’)
(e’)
(
a)
(
c)
(
b)
(c)
Figure 6. ͑a͒ Diffraction-contrast bright-field image, ͑b͒ selected area dif-
fraction pattern of the shell region, and ͑c͒ selected area diffraction pattern of
the core region of Li1.32Ni0.25Mn0.75O2.41 sample.
(
d)
e)
our continuing research is to understand these differences. A detailed
microstructural and chemical investigation using TEM and associ-
ated spectrometry is in progress and will be reported elsewhere.
(
Electrochemical characteristics.— Figure 7 shows the voltage
profiles of cells constructed with cathodes of samples ͑a͒–͑e͒. The
cells were galvanostatically charged and discharged according to the
following procedure: initial charge from the open-circuit voltage to
−
1
4
.9 V and then discharge to 2 V under 23 mA g current density
1
0 µm
1 µm
͑first cycle͒; further charge and discharge between 2 and 4.6 V for
three cycles under the same current density ͑cycles 2–4͒; and within
the same voltage window, cycle three times while doubling the cur-
rent density to 46 mA g ͑cycles 5–7͒, which was repeated until
the current density reached 230 mA g or 10 times the initial
value ͑cycles 11–13͒. We prefer to use the mass current density
instead of the C-rate because of the confusion that the latter can
create among researchers.
Figure 4. SEM images ͑left side: secondary particles and right side: primary
particles͒ of ͑a͒ Li1.32Ni0.25Mn0.75O2.41, ͑b͒ Li1.375Ni0.25Mn0.75O2.4375, ͑c͒
−1
−1
Li1.5Ni0.25Mn0.75O2.5
Li1.65Ni0.25Mn0.75O2.575
,
͑d͒
Li1.575Ni0.25Mn0.75O2.5375
,
and
͑e͒
.
Among all studied materials, the electrochemistry of cathodes
with samples ͑a͒ and ͑b͒ has unique characteristics. First, cells with
these cathodes showed the largest charge and discharge capacities
under all rates, with sample ͑b͒ delivering the highest discharge
growth direction ͑Fig. 6c͒. Owing to the observation of the core/
¯
shell-type morphology, the spinel ͑Fd3m͒ phase is likely physically
integrated within the particles of samples ͑a͒ and ͑b͒. Samples ͑d͒
and ͑e͒, whose secondary particles are constructed with bigger pri-
mary particles, did not show the core/shell morphology. The aim of
−
1
capacity under 230 mA g current density ͑Fig. 7a and b͒. Indeed,
its 210 mAh g− capacity is the highest achieved discharge capac-
ity under such a high current density for an advanced cathode ma-
terial, in addition to the benefit of being cobalt- and
1
4
,5,12,15-17
coating-free.
Moreover, both cathodes exhibited a revers-
ible discharge and charge behavior at around 3.15 V, starting from
the first discharge, which is absent in cathodes with samples ͑c͒–͑e͒
Shell
Core
͑
see Fig. 7aЈ–eЈ͒. In relation to their structural characteristics, cath-
odes of samples ͑a͒ and ͑b͒, whose XRD patterns clearly showed the
existence of a third phase, are distinguished by possessing an addi-
tional lithium intercalation phenomenon starting from the first dis-
charge. This third phase may play some role in the observed elec-
trochemical behavior. In fact, the precursor Ni0.25Mn0.75CO used to
3
prepare all samples is also prone to generate the spinel
LiNi0.5Mn1.5O4 ͑Li0.5Ni0.25Mn0.75O ͒ in much lower lithium con-
2
tent. The highest X-ray peaks ͓͑111͒, ͑113͒, and ͑004͔͒ of this spinel
¯
͑
Fd3m͒ phase overlap with the highest peaks ͓͑003͒, ͑101͒, and
10 µm
¯
͑
104͔͒ of the layered ͑R3m͒ phase, with the possibility of better
separation for the ͑113͒ and ͑004͒ peaks observed at higher angles
͑Fig. 2͒. Because left shoulders appeared in the diffraction lines
͓͑101͒ and ͑104͔͒ of samples ͑a͒ and ͑b͒, we attributed their unique
Figure 5. Backscattered electron image of Li1.32Ni0.25Mn0.75O2.41 ͑sample
a͔͒.
͑
Downloaded on 2014-12-21 to IP 130.209.6.50 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, 157 ͑4͒ A447-A452 ͑2010͒
A451
5
4
3
3
00
nd
1
st charge
2
charge 1st charge
Charge
Discharge
Li1.32Ni0.25Mn0.75O2.41
250
1
st discharge
2
00
50
(
a)
(a')
(b')
300
250
2
5
1
2
1
1
00
50
00
4
3
Li1.375Ni0.25Mn0.75O2.4375
100
5
0
0
5
0
0
10
20
30
40
50
(
b)
Cycle number
2
5
0
0
40
80
120
160
200
Li1.5Ni0.25Mn0.75O2.5
4
3
Cycle number
Figure
8.
͑Color
online͒
Cycling
performance − of
1
Li Ti O /Li1.375Ni0.25Mn0.75O2.4375 ͓sample ͑b͔͒ cell under 73 mA g cur-
(
c')
4
5
12
(
c)
rent density. The inset is the capacity of a Li/Li1.375Ni0.25Mn0.75O
cell vs
2.4375
2
5
cycle number under the same current density.
Li1.575Ni0.25Mn0.75O2.5375
4
3
tured Li1.375Ni0.25Mn0.75O2.4375͓͑sample ͑b͔͒ in a cathode-limited
cell configuration with Li Ti O as the negative electrode. In this
(
d')
4
5
12
(
d)
2
5
case, Li Ti O was selected as the anode to eliminate the possibil-
4 5 12
5
0
st
-1
ity of cycle life fade that could arise from the use of metallic lithium
or graphite. As shown in the inset of Fig. 8, the capacity of a
Li/Li1.375Ni0.25Mn0.75O2.4375 cell slowly degraded with time. The
Li Ti O /Li1.375Ni0.25Mn0.75O2.4375 cell was cycled between 0.5
1
2
5
8
1
cyc.(23 mA.g
cyc.(23 mA.g
cyc.(46 mA.g
)
nd
th
th
-1
st
)
1 cyc.
-1
)
nd
4
3
2
-1
2
cyc.
cyc.(92 mA.g ) -1
th
1
cyc.(230 mA.g
)
4
5
12
−1
Li1.65Ni0.25Mn0.75O2.575
and 3.1 V under a current density of 73 mA g . The cell showed
excellent cycling performance and more than 220 mAh g capac-
ity for at least 200 cycles with very high efficiency.
−
1
(e')
(
e)
0
50 100 150 200 250 300 350 2.0
2.5
3.0
3.5
4.0
4.5
5.0
We also investigated the safety characteristics of the
Li1.375Ni0.25Mn0.75O2.4375 material. The DSC results are shown in
-1
Capacity, mAh.g
Voltage, V
Fig. 9. In the presence of 1.2 M LiPF –EC–EMC electrolyte, the
charged material exhibited a single exothermal peak with an onset
6
Figure 7. ͑Color online͒ Voltage profiles ͑left side͒ at increased current rates
of
͑a͒
Li1.32Ni0.25Mn0.75O2.41 Li1.375Ni0.25Mn0.75O2.4375
,
͑b͒
,
͑c͒
͑e͒
temperature of 275°C. The total amount of heat generated during the
Li1.5Ni0.25Mn0.75O2.5
,
͑d͒ , and
Li1.575Ni0.25Mn0.75O2.5375
−1
thermal decomposition of the material was 830 J g . For compari-
Li1.65Ni0.25Mn0.75O2.575; corresponding differential capacities vs voltage
right side͒.
son, the DSC curve of the commercially available material fully
charged LiNi0.8Co0.15Al0.05O2 is shown in Fig. 9. The high onset
temperature and low heat associated with the electrolyte and charged
͑
¯
electrochemical behavior to the chemically feasible phase ͑Fd3m͒.
Furthermore, when the spinel LiNi0.5Mn1.5O is charged to 4.9 V,
4
2
+
35
most of its charge capacity that is due to the oxidation of Ni ions
4
+
18
to Ni ions should arise at around 4.7 V. However, because the
3
2
2
0
5
0
¯
Fd3m phase is a minority component, its capacity is contained in the
Li1.375Ni0.25Mn0.75O2.4375
¯
capacity of the majority layered components ͑R3m and C2/m͒ dur-
-1
∆H = 830 J.g
ing the initial charging to 4.9 V. Because the latter do not show any
electrochemical activity below 3.5 V, the reversible capacity ob-
served in the cathodes of samples ͑a͒ and ͑b͒, around 3.15 V, starting
15
10
5
LiNi0.8Co0.15Al O2
0.05
4
+
from the first discharge, may mainly be due to the reduction of Mn
-1
18,21
∆H = 1880 J.g
3
+
ions to Mn in the spinel LiNi0.5Mn1.5O .
This hypothesis is
4
further supported by the absence of the capacity region around 3.15
¯
V in the cathodes of samples ͑c͒–͑e͒, where the layered phases ͑R3m
and C2/m͒ were the only components detected in the XRD patterns
0
of these samples ͑Fig. 2͒.
-
5
In addition, samples ͑a͒ and ͑b͒ exhibit nanoprimary particles,
which favor high rate capability because of the short lithium diffu-
sion pathway. These two samples also contain a small amount of a
spinel phase, which is known to yield excellent rate capability, and
thus may contribute to the overall improvement of the rate capabil-
ity.
1
00
150
200
250
300
350
o
Temperature, C
Figure 9.
͑Color online͒ DSC curve of fully charged
Li1.375Ni0.25Mn0.75O2.4375 ͓sample ͑b͔͒ and LiNi0.8Co0.15Al0.05O2 in the pres-
Figure 8 shows the cycling performance of the nanosized struc-
ence of 1.2 M LiPF –EC–EMC electrolyte.
6
Downloaded on 2014-12-21 to IP 130.209.6.50 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A452
Journal of The Electrochemical Society, 157 ͑4͒ A447-A452 ͑2010͒
Li1.375−xNi0.25Mn0.75O2.4375 clearly confirm that this material has
suitable thermal stability for practical cell utilization, especially for
PHEVs and EVs.
References
1
. T. Ohzuku and Y. Makimura, Chem. Lett., 2001, 642.
2. Z. Lu, D. D. MacNeil, and J. R. Dahn, Electrochem. Solid-State Lett., 4, A191
2001͒.
. S. S. Shin, Y.-K. Sun, and K. Amine, J. Power Sources, 112, 634 ͑2002͒.
. S.-H. Kang, Y.-K. Sun, and K. Amine, Electrochem. Solid-State Lett., 6, A183
͑
Conclusions
3
4
Of the five Li͑1+x͒Ni0.25Mn0.75
O
͑0.32 Յ x Յ 0.65͒ com-
͑2.25+x/2͒
͑
2003͒.
pounds investigated, nanostructured ͑a͒ Li1.32Ni0.25Mn0.75O2.41 and
b͒ Li1.375Ni0.25Mn0.75O2.4375 with less lithium compared to
5. Z. Lu and J. R. Dahn, J. Electrochem. Soc., 149, A815 ͑2002͒.
6. A. R. Armstrong, A. D. Robertson, and R. G. Bruce, J. Power Sources, 146, 275
͑
,
͑2005͒.
Li1.5Ni0.25Mn0.75O2.5 ͓or Li͑Li0.2Ni0.2Mn0.6͒O ͔, exhibited the best
2
7
. J. Bréger, M. Jiang, N. Dupré, Y. S. Meng, Y. S. Horn, G. Ceder, and C. P. Grey,
electrochemical performance. All five compositions were synthe-
J. Solid State Chem., 178, 2575 ͑2005͒.
8. W. S. Yoon, N. Kim, X. Q. Yang, J. McBreen, and C. P. Grey, J. Power Sources,
119–121, 649 ͑2003͒.
sized using the Ni0.25Mn0.75CO carbonate precursor made by a co-
3
precipitation method in a continuously stirred tank reactor. With an
increase in the lithium content, the material underwent dramatic
changes structurally, morphologically, and electrochemically.
Among all the studied materials, the samples with the lower lithium
content ͓͑a͒ and ͑b͔͒ exhibited ͑i͒ high capacity at high rate capabil-
9
. C. S. Johnson, J. S. Kim, C. Lefief, N. Li, J. T. Vaughey, and M. M. Thackeray,
Electrochem. Commun., 6, 1085 ͑2004͒.
0. C. H. Lei, J. Bareno, J. G. Wen, I. Petrov, S.-H. Kang, and D. P. Abraham, J. Power
Sources, 178, 422 ͑2008͒.
1
1
1. A. R. Armstrong, M. Holzapfel, P. Novák, C. S. Johnson, S.-K. Kang, M. M.
−
3
ity, ͑ii͒ high packing densities ͑2.1 g cm ͒ that can lead to higher
volumetric energy density, and ͑iii͒ improved safety characteristics.
This excellent performance can be attributed to the existence of a
spinel phase and the dense nanoprimary structure. It is anticipated
that this material could be a potential candidate for the application
of rechargeable Li-ion batteries in PHEVs.
Thackeray, and P. G. Bruce, J. Am. Chem. Soc., 128, 8694 ͑2006͒.
2. S. H. Kang and K. Amine, J. Power Sources, 146, 654 ͑2005͒.
3. S. W. Oh, S. T. Myung, H. J. Bang, C. S. Yoon, K. Amine, and Y.-K. Sun, Elec-
trochem. Solid-State Lett., 12, A181 ͑2009͒.
4. A. D. Robertson and P. G. Bruce, Electrochem. Solid-State Lett., 7, A294 ͑2004͒.
5. J. H. Kim, C. W. Park, and Y.-K. Sun, Solid State Ionics, 164, 43 ͑2003͒.
1
1
1
1
16. Y. J. Kang, J.-H. Kim, S.-W. Lee, and Y.-K. Sun, Electrochim. Acta, 50, 4784
2005͒.
͑
Acknowledgments
17. J. M. Zheng, Z. R. Zhang, X. B. Wu, Z. X. Dong, Z. Zhu, and Y. Yang, J. Elec-
trochem. Soc., 155, A775 ͑2008͒.
This research was funded by the U.S. Department of Energy,
FreedomCAR, and Vehicle Technologies Office. Argonne National
Laboratory is operated for the U.S. Department of Energy by UChi-
cago Argonne, LLC, under contract DE-ACOZ-06CH11357.
1
8. H. Deng, I. Belharouak, Y.-K. Sun, and K. Amine, J. Mater. Chem., 19, 4510
2009͒.
͑
19. Z. Lu, L. Y. Beaulieu, R. A. Donaberger, C. L. Thomas, and J. R. Dahn, J. Elec-
trochem. Soc., 149, A778 ͑2002͒.
2
0. S.-H. Kang and M. M. Thackeray, Electrochem. Commun., 11, 748 ͑2009͒.
Argonne National Laboratory assisted in meeting the publication costs of
this article.
21. S.-H. Park, S.-H. Kang, C. S. Johnson, K. Amine, and M. M. Thackeray, Electro-
chem. Commun., 9, 262 ͑2007͒.
Downloaded on 2014-12-21 to IP 130.209.6.50 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).