A1406 Journal of The Electrochemical Society, 160 (9) A1406-A1414 (2013)

0013-4651/2013/160(9)/A1406/9/$31.00 © The Electrochemical Society

Effect of Structural Integration on Electrochemical Properties of

0.5Li2 MnO3 -0.5Li(Mn0.375 Ni0.375 Co0.25 )O2 Composite Cathodes
for Lithium Rechargeable Batteries
C. Ghanty,a S. Chatterjee,b R. N. Basu,c and S. B. Majumdera,z
a Materials Science Centre, Indian Institute of Technology, Kharagpur 721302, India
b Department of Chemistry, Indian Institute of Technology, Guwahati 781039, India
c Fuel Cell and Battery Division, CSIR-Central Glass and Ceramic Research Institute, Kolkata 700 032, India

In the present work, we have investigated the effect of the structural integration between Li2 MnO3 and Li(Mn0.375 Ni0.375 Co0.25 )O2
components on the electrochemical properties of the resultant compounds. We have adopted three processing methodologies to
attain the structural integration of Li2 MnO3 component in Li(Mn0.375 Ni0.375 Co0.025 )O2 layered cathode. The best electrochemical
properties are achieved in the composites where Li2 MnO3 type nano-domains were formed in situ in Li(Mn0.375 Ni0.375 Co0.25 )O2
particles. For these composites the 1st cycle discharge capacity (at 10 mAg−1 rate) was measured to be 300 mAhg−1 with a
capacity retention ∼220 mAhg−1 after 50 cycles. These cathodes exhibit decent rate capability with typical discharge capacity ∼120
mAhg−1 at 300 mAg−1 rate. One of the remarkable finding of the present work is to demonstrate that in order to yield attractive
electrochemical performance of such integrated cathode; structural integration between the Li2 MnO3 and Li(Mn0.375 Ni0.375 Co0.25 )O2
is essential. To achieve better structural integrity in the physically mixed composites, finer particle size of the individual (Li2 MnO3
and Li(Mn0.375 Ni0.375 Co0.25 )O2) components, through intermixing between them, and optimization of calcinations temperature and
time are recommended.
© 2013 The Electrochemical Society. [DOI: 10.1149/2.040309jes] All rights reserved.

Manuscript submitted April 16, 2013; revised manuscript received June 4, 2013. Published June 26, 2013.

Presently lithium ion rechargeable batteries are the state of the art obtained in 0.5Li2 MnO3 -0.5Li(Mn0.375 Ni0.375 Co0.25 )O2 composition.
power sources for portable electronic devices with relatively lower As recently pointed out in some literature reports, it remains very
power requirements. To meet the current need of various storage difficult to confirm the exact stiochiometric yield of Li2 MnO3 and
devices, rechargeable batteries with higher energy and power den- Li(Mn0.375 Ni0.375 Co0.25 )O2 contents in such composite.4,5 Since
sities are in great demand. To develop batteries with higher energy the diffraction peaks of Li2 MnO3 are almost superimposed on
and power densities novel high voltage, high capacity cathode ma- the layered oxide peaks, quantitative phase estimation through
terials need to be developed. The discharge capacities of traditional X-ray diffraction analysis is difficult. The presence of Li2 MnO3
hexagonal layered oxide, cubic spinel, as well as cathode materials type domains in layered oxide matrix are identified through high
with olivine structure are limited maximum to about 200 mAhg−1 . resolution transmission electron microscopy.6–8 Since, it is not
The nominal voltages of these cathodes are about 3.6 V. Lithium guaranteed that these domains are uniformly dispersed in these
rich Li(Lix Mna Nib Coz )O2 layered oxide based cathode materials are composites, the relative phase fraction could not be estimated
found to be attractive due to their higher capacity (in the range of 250– unambiguously from the HRTEM image. In order to investigate the
280 mAhg−1 ) in the operative voltage window between 4.8–2.0 V.1,2 effect of Li2 MnO3 content on the electrochemical properties of such
Unlike conventional cathode materials, in these lithium rich cath- integrated cathodes, recently effort has been made to alloy Li2 MnO3
odes, lithium ions are also present in the transition metal site. These mechano-chemically with a variety of layered oxide cathode
lithium ions make super-lattice ordering with the manganese ions to materials (including LiNi0.5 Co0.2 Mn0.3 O2 , LiMn0.33 Ni0.33 Co0.33 O2
form Li2 MnO3 type domains.2 Usually these Li2 MnO3 type domains etc) in different proportions.4,5 These mechano-chemically alloyed
are electrochemically inactive; however, it is well known that dur- composites have also been calcined at 1000◦ C for 1 h to investigate
ing charging beyond 4.5 V, Li ions are extracted from Li2 MnO3 type if there is any structural change of these components. Such high
domain with concomitant release of lattice oxygen to form activated temperature heat-treatment affects the electrochemical properties of
oxide. Upon discharge, lithium ions are intercalated into activated these alloys. We feel that mechano-chemical alloying, followed by
manganese oxide and Mn ions are reduced to +3 states. These man- high temperature heat-treatment may not be an effective processing
ganese ions are oxidized to +4 valence state in subsequent charging methodology to understand the nature of structural integration of
cycle. The capacity is increased due to excess lithium ion in tran- Li2 MnO3 in layered oxide cathode. This is due to the fact that high
sition metal site and occurrence of Mn3+ /Mn4+ redox couple along energy ball mill only reduces the particle size of the individual parti-
with usual Ni2+ /Ni4+ and Co3+ /Co4+ redox couples present. How- cles with minimal structural integration.5 The structural integration
ever, in spite of its higher capacity, these cathodes suffer from poor between the Li2 MnO3 and LiMO2 (M = Mn, Ni, Co) type layered
rate capability and capacity fading with repeated charge-discharge oxide components only takes place during high temperature calcina-
cycling.1 tion treatment with the substantial growth of secondary particles.5
For illustrative purpose Li(Lix Mna Nib Coc )O2 type of cathodes Additionally, we feel that the high energy ball mill treatment is too
have been represented as xLi2 MnO3 -(1-x)Li(Mnp Niq Cor )O2 harsh to introduce structural disorder into the constituent cathode
composite. For example, cathodes with composition particles.
Li(Li0.2 Mn0.54 Ni0.13 Co0.13 )O2 may be represented as 0.5Li2 MnO3 - In the present work we have adopted three processing
0.5Li(Mn0.333 Ni0.333 Co0.333 )O2 in composite notation. In a recent methodologies to structurally integrate Li2 MnO3 component into
paper we have demonstrated that the electrochemical prop- Li(Mn0.375 Ni0.375 Co0.25 )O2 layered cathode. Our aim here is to inves-
erties of such lithium and manganese rich compositions are tigate the nature of structural integration achieved through these pro-
strongly dependent on the relative phase fraction of Li2 MnO3 and cesses and its effect on the electrochemical properties of the resultant
Li(Mn0.375 Ni0.375 Co0.025 )O2 in the composite cathode.3 Thus, in composite cathode. Accordingly, we have synthesized 0.5Li2 MnO3 -
case of xLi2 MnO3 -(1-x)Li(Mn0.375 Ni0.375 Co0.25 )O2 (0.0 ≤ × ≤ 1.0) 0.5Li(Mn0.375 Ni0.375 Co0.25 )O2 composite cathodes by varying the na-
composite series, the best electrochemical characteristics are ture of structural integration between the two constituent. Finally we
have proposed a model to the explain dependence of the electrochem-
ical properties on the nature of structural integration of Li2 MnO3
and Li(Mn0.375 Ni0.375 Co0.25 )O2 components in these composite
z
E-mail: [email protected] cathodes.
Downloaded on 2014-12-09 to IP 203.110.246.22 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, 160 (9) A1406-A1414 (2013) A1407

Experimental thick. Finally, disks of 15 mm diameter were punched out (MSK-T06

MTI Corporation, USA) from the laminates and used as cathodes.
Using auto combustion synthesis we have prepared Lithium metal foils were used as counter and reference electrodes.
Li1.0 (Li0.20 Mn0.55 Ni0.15 Co0.1 )O2 (0.5Li2 MnO3 -0.5Li(Mn0.375 Ni0.375 LiPF6 (1M) in a 1:1 (by weight) mixture of ethylene carbonate and
Co0.25 )O2 in composite notation) composite cathodes with vary- diethyl carbonate was used as electrolyte. Celgard 2400, soaked with
ing degree of structural integration. In the first methodology, electrolyte was used as a separator between the anode and cathode.
active (Li(Mn0.375 Ni0.375 Co0.025 )O2 , abbreviated as A) and inactive To fabricate CR2032 type coin cell, the cathode, separator, and an-
(Li2 MnO3 , abbreviated as I) cathode particles are prepared sep- ode were crimped together using a crimping machine (MT-190–20,
arately by auto-combustion synthesis. To prepare these powders, MTI Corporation, USA) kept inside a moisture and oxygen controlled
stoichiometric amount of the acetate salts of lithium, manganese (<1.0 ppm) glove box (Labstar, Mbraun Inc, Germany). Galvanos-
(for Li2 MnO3 ), or lithium, manganese, nickel, and cobalt (for tatic charge-discharge cycling in the voltage range 2.5 and 4.8 V was
Li(Mn0.375 Ni0.375 Co0.025 )O2 ) are dissolved separately in warm carried out with an automated battery tester (BST8-MA, MTI Corpo-
distilled water (∼60◦ C). The salt solutions were mixed together with ration, USA). A potentiostat-galvanostat system (G-750, Gamry Inc.
continuous stirring for 15 min. The solution, thus prepared, was USA) and allied electrochemical analysis software (PHE 200 and EIS
mixed with citric acid solution maintaining citrate to metal ion ratio 300) were used for cyclic voltammetry, and impedance spectroscopy
∼1:1. The solution up on heating at ∼120◦ C (with constant stirring) measurements.
was turned into a viscous gel. Up on further heating the gel was auto-
ignited to form a black fluffy powder. The as prepared powders were
heated at 450◦ C for 3 h for organic removal followed by calcining Results
them at 800◦ C for 12 h for crystallization. The calcined Li2 MnO3 and Structural and micro-structural characteristics.— Powder XRD
Li(Mn0.375 Ni0.375 Co0.025 )O2 powders were wet ball-milled together profiles of the samples active (A), inactive (I), heat and mixed (HM),
for 300 minutes, dried, and used to prepare cathode laminates. This mixed and heat (MH), and chemically mixed (CM) cathode powders
processing methodology is termed as “heat and mixed” abbreviated as are shown in Fig. 1a. The XRD pattern of Li(Mn0.375 Ni0.375 Co0.25 )O2
HM type composite. In the second methodology, similar to that done (A) and Li2 MnO3 (I) are indexed as space group R3̄m (rhombohedral)
for HM composites, Li2 MnO3 and Li(Mn0.375 Ni0.375 Co0.025 )O2 cath- and C2/m (monoclinic) respectively. In line to the recent research re-
ode particles are prepared separately by auto-combustion synthesis. ports by Amlaraj et al.10 and Yu et al.,11 we have also found it appropri-
The as prepared powders were calcined at 450◦ C for 3 h for organic ate to interpret the XRD pattern of the composite powders (MH, HM
removal and then wet ball-milled for 300 minutes for homogeneous and CM of composition 0.5Li2 MnO3 -0.5Li(Mn0.375 Ni0.375 Co0.25 )O2 )
mixing. After drying, the mixed powders were calcined at 800◦ C as two phase structures. Thus, the diffractograms of these powders are
for 12 h for structural integration. This processing methodology is indexed as the superimposed diffraction lines from both rhombohedral
termed as “mixed and heat” abbreviated as MH type composite. In (R) and monoclinic phase (M). Comparing the X-ray diffractograms
the third methodology, stoichiometric amount of the acetate salts of of pure A and I powders we have found only marginal difference in
lithium, manganese, nickel, and cobalt are dissolved separately in the 2θ position of the (201)M (105)R ; (132)M (107)R ; (202)M (060)M ;
warm distilled water (∼60◦ C). The salt solutions were mixed together (108)R (110)R ; (330)M (113)R diffraction peaks. The other lower angle
with continuous stirring for 15 min. The solution thus prepared was diffraction peaks of both A and I phases appears almost in identical
mixed with citric acid solution maintaining citrate to metal ion ratio positions. The relatively lower intensity peaks in the 2θ range ∼20–
∼1:1. The solution up on heating at ∼120◦ C (with constant stirring) 25◦ are due to super-lattice ordering of lithium and manganese in
was turned into a viscous gel. Up on further heating the gel was transition metal sites. In Fig. 1b, we have plotted the slow-scanned
auto-ignited to form a black fluffy powder. The resultant powder XRD peaks of A, I, HM, MH, and CM powders in the 2θ range
was calcined first at 450◦ C for 3h for organic removal followed by
calcination at 800◦ C for 12 h in air for crystallization. This processing
methodology is termed as “chemical mixing” abbreviated as CM
type composite. During the calcination of the auto-combustion syn-
thesized cathode powders, lithium and manganese ions are expected
to order in situ to form Li2 MnO3 type nano-domains embedded
in Li(Mn0.375 Ni0.375 Co0.025 )O2 particle. For such composite, we do
not have any control on the size and distribution of such Li2 MnO3
nano-domains. Additionally, relative phase fraction of Li2 MnO3 type
domain in Li(Mn0.375 Ni0.375 Co0.025 )O2 matrix is difficult to estimate.
The phase formation behavior of the calcined powder is charac-
terized using an X-ray diffractometer (Ultima III, Rigaku Inc. Japan).
X-ray diffractograms of the cathode powders are recorded at a scan-
ning speed of 1◦ min−1 in 2θ range from 10 to 70◦ using CuKα
radiation. The XRD data has been refined by Rietveld method using
MAUD software.9 The surface morphology of the synthesized cathode
powders were characterized using a field emission scanning electron
microscope (SUPRA-40, Carl Zeiss SMT, Germany). The high res-
olution transmission electron micrographs (HRTEM) of the cathode
particles are characterized using a transmission electron microscope
(JEM 2100, JEOL Ltd. Japan).
The electrochemical characteristics of the synthesized cathode
powders were investigated by fabricating CR2032 type coin cells.
To fabricate the cathode composite, a slurry was prepared by mix-
ing cathode powders with acetylene black (as conducting agent) and
polyvinylidene fluoride (PVDF) (as binder) (in a weight ratio of
80:8:12 respectively) in NMP solvent. The slurry was tape cast on Figure 1. (a) X-ray diffractograms of Li(Mn0.375 Ni0.375 Co0.25 )O2 (A),
aluminum foil. The tape cast layer was vacuum dried overnight at Li2 MnO3 (I), and heat and mix (HM), mixed and heat (MH) and chemi-
100◦ C. The dried cathode laminates were then roll pressed (MR- cally mixed (CM) composite powder. (b) Slow scanned XRD profiles of A, I,
500, MTI Corporation, USA) to form cathode layers about ∼60 μm HM, MH, and CM in the two theta range 62–70◦ .

Downloaded on 2014-12-09 to IP 203.110.246.22 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 A1408 Journal of The Electrochemical Society, 160 (9) A1406-A1414 (2013)

shown in Table I. The value of σ (goodness of the fit) and RB (Bragg

factor) are found to be ∼1.4 and 28% respectively. These values in-
dicate a reasonably good quality structural refinement.10,11 Similar
refinements have also been performed for other (active, inactive, and
HM and CM type composites) cathode compositions (not shown).
The refined lattice parameters for all these cathodes are tabulated in
Table II. Inspecting Table II it is apparent that the lattice parameters
of the inactive and active phases in HM composites are marginally
changed as compared to pure active and inactive components. In con-
trast, the inactive phase lattice parameter is grossly changed both in
MH and CM type composites as compared to pure inactive phase.
Field emission scanning electron micrograph of A, I, HM,
MH, and CM are shown in Fig. 3a–3e. The average particle size
of Li2 MnO3 (I) (Fig. 3b) is 4 to 5 times larger than that of
Li(Mn0.375 Ni0.375 Co0.25 )O2 (A) particles (Fig. 3a. For HM composite,
the relatively larger particles are Li2 MnO3 , which are easily distin-
guishable from the relatively smaller Li(Mn0.375 Ni0.375 Co0.025 )O2 par-
ticles (Fig. 3c). As shown in Fig. 3d, for MH type composite, Li2 MnO3
and Li(Mn0.375 Ni0.375 Co0.025 )O2 particles are not easily distinguish-
able. It indicates that in this type of composite, particle growth of
Figure 2. The experimental, refined and residual XRD pattern of MH type Li2 MnO3 is inhibited. Since individual powders are thoroughly mixed
composite cathodes. and calcined at 800◦ C for 12 h, as envisaged in Fig. 3d, they are sin-
tered together to form agglomerated particles. In these agglomerated
particles the particle connectivity is increased. The surface morphol-
62–70◦ . Note that unlike MH and CM composites, in HM compos- ogy of CM composite particle (shown in Fig. 3e) looks similar to that
ites the (202)M (060)M ; (108)R (110)R diffraction peaks can easily be observed for MH type composite (Fig. 3d). Note that in accordance to
identified. Additionally, (113)R and (330)M peaks are also identified our understanding, in CM type composite nano-domains of Li2 MnO3
separately in HM composite. It is inferred that in these composites, forms in situ in Li(Mn0.375 Ni0.375 Co0.025 )O2 type matrix.3,6–8
Li2 MnO3 and Li(Mn0.375 Ni0.375 Co0.25 )O2 particles stay separated. In The nature of the structural integration of I and A phases both in
contrast to HM, in CM and MH composites, structure integration of MH and CM type composites are further investigated using transmis-
I and A phases have taken place (presumably with varying degree) sion electron microscopy. Fig. 4a and 4b compares the morphology
during calcinations treatment. of MH and CM composite powders. Since after organic removal at
To better understand the nature of such structural integration the relatively lower temperature (∼450◦ C) both these powders are cal-
XRD patterns of active, inactive as well as HM, MH, and CM type cined at higher temperature (∼850◦ C) for elongated time (∼12 h),
composites are refined using Rietveld refinement. For refinement of most of the particles are partially sintered to form agglomerates. A se-
these composites we have used similar model used by F. Amalraj ries of high resolution transmission electron micrographs are recorded
et. al.10 and Yu et al.11 According to their model all the XRD peaks (with progressively higher magnification) in the regions marked in
of these integrated cathodes are assumed to be the superimposed Fig. 4a and 4b and these micrographs are shown in Fig. 4c (for MH
diffraction lines from rhombohedral Li(Mn0.375 Ni0.375 Co0.25 )O2 and type composite) and Fig. 4d (for CM type composite) respectively.
monoclinic Li2 MnO3 phases. As a typical example, for MH com- Comparing the HRTEM micrographs of MH and CM type compos-
posite, the refined, experimental and residual pattern is shown in ites, (taken at highest magnification) it is concluded that in MH type
Fig. 2. The atomic coordinates, structural and statistical parameters, composite, Li2 MnO3 and Li(Mn0.375 Ni0.375 Co0.025 )O2 particles are sin-
estimated from the X-ray Rietveld refinement of MH composite, is tered together with a interface region formed during the calcinations.

Table I. Atomic coordinates, structural and statistical parameters estimated from Rietveld refinement of MH composite.

MH: Phase 1: Li(Mn0.375 Ni0.375 Co0.25 )O2 Phase 2: Li2 MnO3 Rwp = 40%, RB = 28%, Rexp = 28%, σ = 1.4
Atomic Coordinate

phase Space group Site Atom x y z Biso Occupancy Density(g/cc) Weight fractions

1 R3̄m 3b Li 0 0 0.5 0.2010 1 4.78 41%

3a Mn 0 0 0 0.3675 0.375
3a Ni 0 0 0 0.3675 0.375
3a Co 0 0 0 0.3675 0.25
6c O 0 0 0.2601 0.7388 1
2 C2/m 4g Mn 0 0.1700 0 0.2144 0.85 3.88 59%
2b Li 0 0.5 0 0.1112 0.504
2b Mn 0 0.5 0.5 0.1112 0.41
2c Li 0 0 0.5 0.5990 1
4h Li 0 0.6998 0.5 0.3639 1
4i O 0.1984 0 0.2127 0.2427 1
8i O 0.2570 0.3052 0.2407 0.6048 1
    
wi |Yi −Yc,i |2 |
hIi,h −Ic,h | n− p Rwp
Rwp = i=1,n
 × 100, R B = × 100, Rex p =  × 100 and σ = Where, Rwp is weighted profile factor, RB is Bragg
i=1,n wi Yi h |Ii.h |
2
i wi Yi
2 Rex p

factor, Rexp is the expected weighted profile factor and σ is the goodness of the fit.9

Downloaded on 2014-12-09 to IP 203.110.246.22 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, 160 (9) A1406-A1414 (2013) A1409

Table II. Comparison of lattice parameters for all the synthesized powders.

samples Active (R3̄m) Inactive (C2/m)

a (Å) c (Å) γ (◦ ) a (Å) b (Å) c (Å) γ (◦ )

A 2.866 (±0.001) 14.252 (±0.001) 120 – – – –

I – – – 4.924 (±0.005) 8.526 (±0.003) 5.091 ±0.009) 109.22
HM 2.865 (±0.002) 14.254 (±0.003) 120 4.927 (±0.02) 8.526 (±0.05) 5.023 (±0.04) 109.07
MH 2.851 (±0.002) 14.255 (±0.002) 120 4.980 (±0.003) 8.541 (±0.002) 5.028 (±0.004) 109.23
CM 2.853 (±0.001) 14.251 (±0.001) 120 4.963 (±0.002) 8.556 (±0.002) 5.019 (±0.001) 109.27

Across the interface region the orientation and spacing of the lattice Through HRTEM analyzes, within individual particles, in situ forma-
fringes are different which signify the dissimilarity between the par- tion of Li2 MnO3 type nano-domain formation is predicted. The elec-
ticles. In contrast to these micrographs, the HRTEM micrographs of trochemical characteristics of Li(Mn0.375 Ni0.375 Co0.025 )O2 , Li2 MnO3
the CM composites exhibit similar type lattice fringe across the in- cathodes, together with CM, MH and HM type composites are de-
terface region of two particles sintered together. However, within an scribed in the subsequent sections.
individual particle, along with the well ordered fringes, there are re-
gions with disturbed lattice fringes (marked in the high magnification Cyclic voltammetry.— For the 1st and 2nd cycles, the cyclic voltam-
HRTEM micrograph, Fig. 4b). These disordered regions correspond mograms (CV) of A, I, HM, MH, and CM cathodes are presented
to the nano-domains of Li2 MnO3 .6–8 The nano-domains (form due to in Fig. 5a–5e. All the CVs were recorded in the voltage window
the ordering of lithium and manganese ions in transition metal layer) 2.5–5.0 V at a scan rate of 0.2 mVs−1 . In accordance to earlier re-
are large enough to yield coherent X-ray diffraction. search reports on similar cathode compositions,12 for the sample A
The results of microstructure evolution can be summarized as (Fig. 5a) the anodic peak at ∼3.95 V is assigned to be due to the
follows: the HM type composites are characterized with separated oxidation of Ni2+ to Ni4+ and Co3+ to Co4+ . The corresponding re-
Li2 MnO3 (relatively bigger in size) and Li(Mn0.375 Ni0.375 Co0.025 )O2 duction peak is observed at ∼3.65 V. The redox peaks around 4.5 V
particles with minimal sintering between these particles. For MH probably appears due to the extraction/ insertion Li ions from/ into the
type composites, particle growth of Li2 MnO3 is grossly retarded, tetrahedral site.13–15 The nature and position of these peaks in both the
and, purely from size variation these particles could not be differ- cycles remains almost same except marginal capacity fading occurs in
entiated from Li(Mn0.375 Ni0.375 Co0.025 )O2 particles. Through HRTEM the second cycle. In the CV of sample I as shown in Fig. 5b, a promi-
analyzes it is concluded that most of the individual Li2 MnO3 and nent oxidation peak ∼4.7 V indicates the extraction of Li+ ions with
Li(Mn0.375 Ni0.375 Co0.025 )O2 are sintered to form agglomerated parti- concomitant release of lattice oxygen. The overall effect of this elec-
cles. Also for CM type composites, agglomeration of individual parti- trochemical reaction is the extraction of Li2 O from Li2 MnO3 to form
cles are observed due to partial sintering between individual particles. activated MnO2 which upon discharge participate in the reversible Li
ion insertion extraction reaction. A small peak around 3.4 V in the first
cathodic sweep is due to the reduction of the Mn4+ ion to Mn3+ ion

Figure 4. Transmission electron micrograph of (a) MH and (b) CM composite

Figure 3. Scanning electron micrographs of (a) A, (b) I, (c) HM, (d) MH, and powders. The high resolution micrograph of (c) MH and (d) CM composite
(e) CM pristine powders. powder from the respective regions marked in (a) and (b).

Downloaded on 2014-12-09 to IP 203.110.246.22 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 A1410 Journal of The Electrochemical Society, 160 (9) A1406-A1414 (2013)

Figure 5. Cyclic voltammograms of (a) A, (b) I, (c) HM, (d) MH and (e) CM cathodes for first two scan cycles.

from activated MnO2 to form LiMnO2 . The formed LiMnO2 is known nature of the CVs of CM and MH are different from that of HM type
to transform electrochemically into a spinel structure. The other peak composite. Thus, as shown in Fig. 5d and 5e, CVs of both MH and CM
below 3.0 V in the first cathodic sweep is due to the insertion of Li ion contain redox peaks for transition metal ion oxidation at ∼4.2 V and
to the octahedral site of the spinel structure. Corresponding oxidation peak for Li2 O extraction ∼4.8 V. Both the samples contain reduction
peaks are also observed in the 2nd anodic sweep around 2.8–3.5 V. peaks corresponding to nickel/cobalt and manganese ions at ∼3.6 and
Interestingly, appearance of anodic peak ∼4.7 V peak in the 2nd scan ∼3.1 V respectively. Interestingly, in the 1st cycle anodic sweep of
suggests that the extraction of Li2 O is not completed in the first cycle MH, there is a shoulder peak at ∼4.0 V along with the ∼4.2 V peak.
and it takes few more cycles to complete the reaction. In the cyclic The splitting of the anodic peak in 4.0 V regions is probably due to the
voltammograms of HM (Fig. 5c), both the anodic peaks for oxidation oxidation of the transition metal ions (Ni and Co) from two different
of transition metals and extraction of Li2 O at ∼4.0 and ∼4.7 V re- environments. The origin of this shoulder peak is discussed in detail
spectively are present along with the characteristics hump at ∼4.5 V latter in the discussion section. For CM type composite cathode, in
similar to that of Li(Mn0.375 Ni0.375 Co0.025 )O2 cathode. The nature of the 1st cycle anodic scan, a peak ∼4.8 V represents Li2 O extraction
CV of HM can be considered simply as the super imposed cyclic from Li2 MnO3 type nano-domain. The signature of electrochemical
voltammograms of the cathode laminates A and I. In contrast, the Li2 O extraction is also apparent in the 2nd anodic scan. The anodic

Downloaded on 2014-12-09 to IP 203.110.246.22 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, 160 (9) A1406-A1414 (2013) A1411

to be 170, 24, 85, 225, and 290 mAhg−1 respectively. Considering the
fact that in HM composite, Li2 MnO3 and Li(Mn0.375 Ni0.375 Co0.025 )O2
molar contents are in equal proportion, yield of the discharge capacity
∼85 mAhg−1 is justified. As compared to Li(Mn0.375 Ni0.375 Co0.025 )O2 ,
Li2 MnO3 , and also HM type composite cathodes, much higher dis-
charge capacities are obtained in MH and CM type cathodes. The
reason for such capacity increase in MH and CM type composite
cathodes are discussed later. It is interesting to note that for I cath-
ode, the discharge capacities in the second cycles is slightly increased
whereas it decreases in A, HM, MH, and CM type cathodes. To better
understand the charge-discharge characteristics of these cathode mate-
rials, we have plotted the differential capacity as a function of voltage
(Fig. 6f–6j). Several interesting features are apparent in these plots.
Similar to that we observed in respective cyclic voltammograms, the
peak and trough ∼3.75 V in A and HM (Fig. 6f and 6h) appear due
to the oxidation and reduction of Ni and Co ions. For HM, the peak
at ∼4.75 V represents the extraction of Li2 O from Li2 MnO3 . Sample
MH and CM also show similar peak ∼4.7 V, but with much larger
intensity. This oxidation peak is absent in the 2nd cycle which indi-
cates that extraction of Li2 MnO3 virtually completed in the first cycle.
The nature of the oxidation peak ∼4.0 V is quite different in CM and
MH than that observed in HM composite. Thus, as shown in Fig. 6i,
for MH composite, there are two distinct oxidation peaks at ∼3.8 V
and ∼4.0 V in the first cycle. These oxidation peaks are merged to a
single peak in 2nd cycle. Also we have noted that the redox peaks for
Ni/Co ions appear ∼3.75 V in the active cathode, however, both for
CM and MH type composites the oxidation peak shifts toward higher
voltage whereas the reduction peak shifts to relatively lower voltage,
as compared to active cathodes.

Cycleability and rate capability characteristics.— Figure 7 shows

the cycling stability of (a) CM and (b) MH composites at different
current densities. In this experiment each cell of particular compo-
sition is cycled galvanostatically between 4.8 to 2.5 V for 50 cycles
at current rates ranging 10–300 mAg−1 . We had always used a fresh
cell for the repeated charge-discharge measurements at each current
density. The used rate profile is as follows: First the cell was charged
up to 4.8 V at constant current (say @100 mAg−1 ) and then kept at the
maximum voltage (4.8 V) till the measured current reduced to 10% of
the charging current (i.e 10 mAg−1 ) (this is known as constant current
constant voltage (CCCV) mode charging). Subsequently the cell was
discharged galvanostatically (say @100 mAg−1 ) to 2.5 V (constant
current (CC) mode discharge). For both CM and MH type composites,
the discharge capacities reduce with the increase of current rate. For
CM type, irrespective of the rate, the discharge capacity is reduced
with repeated charge-discharge cycles. Similar trend is observed in
MH composites at lower rate, however, when the rate is increased to
100 mAg−1 , initially the discharge capacity increases with cycling,
and then it starts to decrease with repeated charge discharge cycling.
To better illustrate the rate capability behavior, the normalized dis-
charge capacity with charge discharge cycling is compared for CM
Figure 6. Charge discharge profiles of ((a) A, (b) I, (c) HM, (d) MH and and MH type composites at 10,100, and 300 mAg−1 current rates (Fig.
(e) CM cathodes. Corresponding differential capacity (dQ/dV vs.V) plots are 7c). Note that at lower rate (up to 100 mAg−1 ), MH type composites
shown in Fig. 6(f)-(j) respectively. yield excellent capacity retention characteristics. However, at higher
rate (∼300 mAg−1 ), we have found that the CM type composite yield
better capacity retention behavior. The energy density of the MH and
and cathodic peaks correspond to nickel/cobalt ions are broadened. In CM type of composite cathodes were estimated as a product of dis-
the 2nd cycle, the anodic and cathodic peaks (∼3.3 V and ∼3.0 V), charge capacity and nominal voltage of the respective cell. For MH
corresponding to the manganese ion oxidation and reduction, are type composite the energy densities were estimated to be ∼800 and
clearly visible. The appearance of these peaks is due to the extraction ∼676 mWhg−1 after 1st and 50th charge-discharge cycles. The energy
and insertion of lithium ion in activated manganese dioxide. density was significantly increased (∼1018 mWhg−1 after 1st cycle)
in CM type composite cathode. For these cathodes the energy density
Charge discharge characteristics.— In the voltage window be- was estimated to be ∼747 mWhg−1 after 50 charge discharge cycles
tween 2.5 to 4.8 V, the galvanostatic charge discharge profiles (at at 10 mAg−1 .
a current rate of 10 mAg−1 ) of the cathode materials prepared As far as the capacity fading of these composites cathodes are
from A, I, HM, MH, and CM composites cathodes are shown in concerned, we feel the nature of electrolyte plays a major role. The
Fig. 6a–6e. The corresponding differential capacity plots are shown high charging voltage of these (∼4.8 V) could trigger an oxidative
in Fig. 6f–6j respectively. As noted in Fig. 6a–6e the first cycle decomposition of the electrolyte,1 which eventually leads to the for-
discharge capacity for A, I, HM, MH, and CM cathodes are measured mation of a solid electrolyte interface (SEI) layer on the surface of

Downloaded on 2014-12-09 to IP 203.110.246.22 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 A1412 Journal of The Electrochemical Society, 160 (9) A1406-A1414 (2013)

Figure 8. Nyquist plot of (a) CM and (b) MH type composite after 10, 25, 40
and 50th cycles (at 50 mAg−1 current). The equivalent circuit model used for
fitting along with a typical Nyquist fit is shown in the inset of (a).

Following this hypothesis, at lower rate, the capacity of CM cathode

should fade more as observed in the present study. For MH type of
composite, at relatively lower discharge rate (up to 50 mAg−1 ), the
capacity fading characteristics are similar to that observed in CM
type composite. However, when discharge current is further increased
(say ∼100 mAhg−1 ), then unlike the CM type composites, oxygen
removal from Li2 MnO3 component is not completed in the 1st charge
cycle and it continues in the subsequent charge cycles. As a result in
first couple of cycles, discharge capacities found to increase progres-
sively (see Fig. 7b and 7c). With further increase of charge-discharge
cycling, the discharge capacity starts to fade due to the increase of
charge transfer resistance.
To better understand the fading characteristics we have performed
electrochemical impedance spectroscopy (EIS) of CM and MS type
Figure 7. Rate capability characteristics of (a) CM and (b) MH composite composites. Fig. 8 shows the Nyquist plots of (a) CM and (b) MS
cathodes. (c) Normalized capacity vs. cycle number plots for these composites. composites after different discharge cycles as marked in the respective
plots. Both in Fig. 8a and 8b, two semi-circles at high and intermediate
frequencies (ω) and a low frequency tail can easily be identified. The
the cathode. Formation of SEI progressively increases the charge high frequency intercept on the real axis denotes the total resistance
transfer resistance leading to capacity fading. For low rate charge- of the electrolyte, separator and electrical contacts (Rs ). The high and
discharge, the electrolyte in the cell is exposed to high voltage for intermediate frequency semi-circles are related to the Li ion migra-
elongated time. Therefore, as compared to high rate charge-discharge tion resistance through solid electrolyte interface (RSEI ) and charge
cycling, the extent of SEI formation should be more at lower rate. transfer resistance (RCT ) in integrated cathode-electrolyte interface

Downloaded on 2014-12-09 to IP 203.110.246.22 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, 160 (9) A1406-A1414 (2013) A1413

Table III. The values of circuit elements estimated from best fitted Nyquist plots.

CPE1(S.s1/2 .cm−2 ) CPE2 (S.s1/2 .cm−2 )

Comp. Cycle RS () RSEI () Q (10−5 ) a RCT () Q (10−3 ) a W (S.s1/2 .cm−2 )
CM 2 8.8 54 2.8 0.7 10 7.9 0.88 0.13
10 8.5 46 4.4 0.65 76 9.2 0.82 0.1
25 6.6 61 5.5 0.57 151 8.0 0.82 0.07
40 5.4 62 3.9 0.57 282 7.5 0.77 0.05
50 4.8 64 2.9 0.58 343 6.9 0.75 0.06
MH 2 4.7 27 1.6 0.78 6 10 0.84 0.14
10 7.3 21 2.3 0.74 37 7.9 0.91 0.12
25 4.1 43 3.4 0.6 407 8.5 0.92 0.04
40 4.9 75 6.1 0.57 406 9.1 0.87 0.02
50 6 82 4.7 0.6 384 7.6 0.78 0.02

respectively. The low frequency tail is associated with the lithium be 225 mAhg−1 for MH type composite and 300 mAhg−1 for CM
ion diffusion into the bulk of the cathode.16–18 We have modeled the type composite. These discharge capacities are far larger as mea-
Nyquist plots to investigate the effect of Rs , RSEI , RCT and the Li+ ion sured either for HM type composite or for individual Li2 MnO3 and
diffusion on the cell resistance. The equivalent circuit model used for Li(Mn0.375 Ni0.375 Co0.025 )O2 cathodes. The capacity of MH and CM
such fitting along with a typical Nyquist fit is shown in the inset of type composites is increased due to the excess lithium ion in transi-
Fig. 8a. All these impedance spectra were recorded in the cathodes tion metal site and occurrence of both Mn3+ /Mn4+ along with usual
cycled at 50 mAg−1 .The values of the fitted circuit elements are also Ni2+ /Ni4+ and Co3+ /Co4+ redox couple in them. Recall that both for
tabulated in Table III. As shown in the Table, for CM composite, the MH and CM type composites, the manganese ion redox reactions are
resistance due to SEI formation (RSEI ) is marginally changed with clearly identified both in cyclic voltammograms (Fig. 5) and differ-
repeated charge-discharge cycling. With cycling, however, there is ential capacity plots (Fig. 6) presented earlier. As Li2 MnO3 particles
a progressive increase of the charge transfer resistance (RCT ) in the are not structurally integrated, such manganese ion redox is not very
CM cathode. The Warburg coefficient of CM composite is reduced prominent in HM type composite. As a result, comparatively poorer
with repeated charge discharge cycling. In contrast, RSEI of the MH discharge capacity is achieved in this composite.
cathode increases with cycling and RCT is dramatically increased after Based on the structural and micro-structural analyzes, the morpho-
10th cycle, and then remains almost constant up to 50th cycles. Also, logical features of the active (A, Li(Mn0.375 Ni0.375 Co0.025 )O2 ), inactive
similar to that observed in CM composite, the Warburg coefficient of (I, Li2 MnO3 ), HM, MH, and CM type composites are shown schemat-
MH composite is reduced with cycling. Reviewing these data we feel ically in Fig. 9. As presented earlier, the inactive component (Fig. 9b)
that the MH processing has its own effect to destabilize the SEI layer. has relatively larger particle size as compared to active component
Interestingly, both for CM and MH type composites, the magnitudes (Fig. 9a). The size of the inactive particle is reduced in MH type
of RSEI and RCT do not differ appreciably after 50th charge-discharge composite (Fig. 9d). The integrated regions of MH type composites
cycles. Probably, due to this reason the fading characteristics for both are created due to the partial sintering of active and inactive particle.
these composites (at 50 mAg−1 ) remain similar in nature. In these regions the chemical environment of lithium and transition
metal cations are different than their environment in active particle.
Discussion In CM type composite (Fig. 9e), nano-size domains of Li2 MnO3 is
structurally integrated in the active particles. Through Rietveld re-
The results presented in the preceding section are summarized finement of the XRD patterns we identified that as compared to pure
as follows: As presented in Fig. 6, the individual discharge ca- Li2 MnO3 and HM composite, the ‘a’ and ‘b’ parameters of Li2 MnO3
pacities of Li2 MnO3 and Li(Mn0.375 Ni0.375 Co0.025 )O2 cathodes af- component in MH and CM composites are increased substantially.
ter 1st discharge is 24 mAhg−1 and 170 mAhg−1 respectively. The observed change in the lattice parameter indicates the structural
Composite prepared by simple mixing of calcined Li2 MnO3 and integration between the active and inactive components in these com-
Li(Mn0.375 Ni0.375 Co0.025 )O2 powders (HM) yield cathode with poor posites. Recall, while explaining the CV and differential capacity plots
discharge capacity ∼85 mAhg−1 . This observation is in line with of MH type composite, we found a broad shoulder (Fig. 5d) and clear
other literature report. Thus, in case of cathode made of hand blended splitting (Fig. 6i) of the oxidation peak in 4 V range. Presumably in
Li2 MnO3 : LiMn0.33 Ni0.33 Co0.33 O2 (1:1) powder, discharge capacity MH type composites, nickel/cobalt ions are in two different chemical
only about 100 mAhg−1 has been reported at C/20 discharge rate.5 environments namely integrated and active regions (see Fig. 9d). Prob-
In HM type composites, through X-ray diffraction in conjunction ably, in these two regions the oxidation potential of nickel and cobalt
with scanning electron micrograph analyzes we have clearly identi- cations is seemingly different which results such broadening (of the
fied isolated Li2 MnO3 and Li(Mn0.375 Ni0.375 Co0.025 )O2 particles with- CV plot) and splitting (in the differential capacity plot). In CM type
out any structural integration. In contrast, by analyzing the lattice
fringe pattern of the interfacial region we had demonstrated a partial
structural integration between monoclinic Li2 MnO3 and rhombohe-
dral Li(Mn0.375 Ni0.375 Co0.025 )O2 phase in the MH type composite. This
kind of structural integration is feasible as both these phase share a
common close packed oxygen lattice.2 The X-ray diffractogram of
the MH type composite is found to have superimposed diffraction
peaks both from monoclinic and rhombohedral phases with C2/m and
R3̄m space groups respectively. For CM type composite, the struc-
tural integration of these monoclinic and hexagonal layered phase
is achieved through the in situ formation of Li2 MnO3 type domain
in Li(Mn0.375 Ni0.375 Co0.025 )O2 particle. In the voltage window be- Figure 9. Schematic of the proposed morphology of (a) active, (b) inactive,
tween 4.8–2.5 V, the 1st cycle discharge capacities are measured to (c) HM, (d) MH and (e) CM type composite cathodes.

Downloaded on 2014-12-09 to IP 203.110.246.22 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 A1414 Journal of The Electrochemical Society, 160 (9) A1406-A1414 (2013)

composites the Li2 MnO3 type domains is embedded into the active yields energy density ∼1018 and ∼747 mWhg−1 at 1st and 50th cycle
matrix and therefore, such oxidation peaks are only broadened and respectively. This cathode exhibit decent rate capability with typical
appear at relatively higher potential as compared to both HM type discharge capacity ∼120 mAhg−1 at 300 mAg−1 rate.
composite and active Li(Mn0.375 Ni0.375 Co0.25 )O2 cathode. One of the remarkable finding of the present work is to demon-
Analyzing the fitted parameters of the Nyquist plots of CM and strate that to yield attractive electrochemical performance of in-
MH type composites, we can infer that the capacity fading of these tegrated cathode, structural integration between the Li2 MnO3 and
cathodes are related to the increase in both RSEI and RCT values with Li(Mn0.375 Ni0.375 Co0.25 )O2 is essential. The structural integration may
cycling. Interestingly, for CM composite RSEI changes marginally with be achieved either by post calcinations of thoroughly mixed pow-
number of cycles, whereas it increases systematically with cycling for ders or through the in situ formation of Li2 MnO3 type domains
MH composites. We feel that MH processing has its own effect in in Li(Mn0.375 Ni0.375 Co0.25 )O2 particles. The nature of structural in-
destabilizing the SEI layer. At this stage, the exact reason for such tegration has been characterized through high resolution transmis-
destabilization is not clear to us however, we feel that it might be re- sion electron microscopy in conjunction with Rietveld refinement of
lated to the progressive evolution of lattice oxygen from the Li2 MnO3 the X-ray diffractograms. As compared to CM type composite, it is
constituent of MH composite. For MH composite, the destabilization easier to estimate the Li2 MnO3 content in MH type composite cath-
of RSEI combined with the sudden rise of the charge transfer resistance odes. Since Li2 MnO3 type nano-domains are fully integrated in active
(RCT ), might be responsible for its poorer cycleability at higher rate. (Li(Mn0.375 Ni0.375 Co0.25 )O2 ) matrix better electrochemical character-
Interestingly, both for CM and MH type composites, the magnitudes istics are achieved in CM type composites. In order to achieve better
of RSEI and RCT do not differ appreciably after 50th charge-discharge structural integration in MH type composites, reduction of the particle
cycles. Probably, due to this reason the fading characteristics for both size of Li2 MnO3 and Li(Mn0.375 Ni0.375 Co0.25 )O2 components, better
these composites (at 50 mAg−1 ) remain similar in nature. Systematic intermixing between them, and optimization of calcinations tempera-
analyzes of the impedance data measured with cycling at various rates ture and time are recommended.
will lead to better understanding on these issues.
Acknowledgment

Conclusions The research work was partially supported by the equipment grant
from the Alexander von Humboldt Foundation, Bonn, Germany. One
In the present work, we have studied the interrela- of the authors, C. Ghanty, would also like to acknowledge CSIR, Govt.
tion between the processing methodology and electrochemi- of India for providing him a Research Fellowship under CSIR-NET
cal characteristics of Li(Li0.2 Mn0.55 Ni0.15 Co0.1 )O2 (0.5Li2 MnO3 - scheme.
0.5Li(Mn0.375 Ni0.375 Co0.25 )O2 , in equivalent composite notation)
integrated cathode materials for high energy density lithium recharge-
able batteries. In the voltage window between 4.8 to 2.5 V, the 1st cycle References
discharge capacities (at 10 mAg−1 rate) of Li(Mn0.375 Ni0.375 Co0.25 )O2 1. R. Marom, S. F. Amalraj, N. Leifer, D. Jacob, and D. Aurbach, J. Mater. Chem., 21,
and Li2 MnO3 cathodes are measured to be 170 mAhg−1 and 9938 (2011).
24 mAhg−1 respectively. Three different type of process methodolo- 2. M. M. Thackeray, C. S. Johnson, J. T. Vaughey, N. Li, and S. A. Hackney, J. Mater.
Chem., 15, 2257 (2005).
gies were adopted to prepare the composite powders. When individu- 3. C. Ghanty, R. N. Basu, and S. B. Majumder, J. Electrochem. Soc., 159, A1125 (2012).
ally calcined Li2 MnO3 and Li(Mn0.375 Ni0.375 Co0.25 )O2 powders were 4. S. Kim, C. Kim, J.-K. Noh, S. Yu, S.-J. Kim, W. C. Choi, K. Y. Chung, and B.-W. Cho,
mixed thoroughly (without any additional calcination treatment to J. Power Sources, 220, 422 (2012).
the mixed powder) the composite cathode yield discharge capacity 5. W. C. West, J. Soler, and B. V. Ratnakumar, J. Power Sources, 204, 200 (2012).
6. J. S. Kim, C. S. Johnson, J. T. Vaughey, M. M. Thackeray, S. A. Hackney, W. Yoon,
∼85 mAhg−1 at 1st cycle. The 1st cycle discharge capacities are and C. P. Grey, Chem. Mater., 16, 1996 (2004).
found to be grossly increased (∼225 mAhg−1 ) when the Li2 MnO3 7. S. Sivaprakash, S. B. Majumder, S. Nieto, and R. S. Katiyar, J. Power Sources, 170,
and Li(Mn0.375 Ni0.375 Co0.25 )O2 powders, after organic removal, were 433 (2007).
thoroughly mixed and calcined at 800◦ C for 12 h in air (MH type 8. S. Sivaprakash and S. B. Majumder, J. Electrochem. Soc., 157, A418 (2010).
9. http://www.ing.unitn.it/˜maud/.
composite). For MH type composite cathodes the capacity retention 10. F. Amalraj, D. Kovacheva, M. Talianker, L. Zeiri, J. Grinblat, N. Leifer, G. Goobes,
was measured ∼195 mAhg−1 after 50 cycles (at 10 mAg−1 ). This B. Markovsky, and D. Aurbach, J. Electrochem. Soc., 157, A1121 (2010).
composition also yields energy density ∼800 and ∼676 mWhg−1 at 11. H. Yu, H. Kim, Y. Wang, P. He, D. Asakura, Y. Nakamura, and H. Zhou, Phys. Chem.
1st and 50th cycle respectively. This cathode exhibit decent rate capa- Chem. Phys., 14, 6584 (2012).
12. X. Li, Y. J. Wei, H. Ehrenberg, F. Du, C. Z. Wang, and G. Chen, Solid State Ionics,
bility with typical discharge capacity ∼120 mAhg−1 at 300 mAg−1 178 1969 (2008).
rate. In the third type of synthesis methodology, Li2 MnO3 type nano- 13. M. S. Whittingham, Chem. Rev., 104, 4271 (2004).
domains were formed in situ in Li(Mn0.375 Ni0.375 Co0.25 )O2 particles 14. M. Jiang, B. Key, Y. S. Meng, and C. P. Grey, Chem. Mater., 21, 2733 (2009).
(CM type composite) and for these composited the 1st cycle discharge 15. F. Wu, M. Wang, Y. Su, S. Chen, and B. Xua, J. Power Sources, 191, 628 (2009).
16. D. Aurbach, J. Power Sources, 89, 206 (2000).
capacity (at 10 mAg−1 rate) was measured to be ∼300 mAhg−1 . For 17. M. G. S. R. Thomas, P. G. Bruce, and J. B. Goodenough, Solid State Ionics, 18–19,
CM type composite cathodes the capacity retention was measured 794 (1986).
∼220 mAhg−1 after 50 cycles (at 10 mAg−1 ). This composition also 18. S. M. Lee, S. H. Oh, W. I. Cho, and H. Jang, Electrochim. Acta, 52, 1507 (2006).

Downloaded on 2014-12-09 to IP 203.110.246.22 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).