You are on page 1of 10

Energy &

Environmental
Science
View Article Online
PAPER View Journal | View Issue

Improving the electrochemical performance of


layered lithium-rich transition-metal oxides by
Cite this: Energy Environ. Sci., 2014, 7,
705 controlling the structural defects
Jinlong Liu, Mengyan Hou, Jin Yi, Shaoshuai Guo, Congxiao Wang and Yongyao Xia*
Published on 11 November 2013. Downloaded on 1/4/2023 6:50:19 PM.

We report the electrochemical properties of layered lithium-rich Li1.2Mn0.54Ni0.13Co0.13O2 cathode


materials with various degrees of stacking faults, which are prepared via a facile molten-salt method
using a variety of fluxes including KCl, Li2CO3, and LiNO3. The frequency of the stacking faults is highly
dependent on the temperature and molten salt type used during the synthesis. A well-crystallized
Received 17th May 2013
Accepted 8th November 2013
Li1.18Mn0.54Ni0.13Co0.13O2 nanomaterial with a larger amount of stacking faults synthesized at 800  C for
10 h in an inactive KCl flux delivers a high reversible capacity of 310 mA h g1 at room temperature,
DOI: 10.1039/c3ee41664j
while the samples prepared in the chemically active fluxes with a smaller amount of stacking faults show
www.rsc.org/ees poor electrochemical performance.

Broader context
Advanced lithium-ion batteries (LIBs) that deliver more energy at rapid charge and discharge rates are essential for on-board storage technology in hybrid
electric vehicles (HEVs) or electric vehicles (EVs). High energy density lithium-rich transition metal oxide cathodes represent an important milestone in
materials design for advanced lithium-ion batteries due to their high reversible capacities of 250 mA h g1 at low cost. Although much progress has been
achieved on the lithium-rich transition-metal oxides, the relationship between the microstructure e.g. stacking faults and electrochemical properties of lithium-
rich transition-metal oxides remains unclear. In this study, we report the electrochemical performance of Li1.2Mn0.54Ni0.13Co0.13O2 electrodes with various
degrees of stacking faults and reveal that structural defects in the crystal structure of the Li2MnO3 component play a key role in the electrochemistry of
xLi2MnO3$(1  x)LiMO2 electrodes using powder X-ray diffraction (XRD), selected area electron diffraction (SAED), Raman spectroscopy, and X-ray photo-
electron spectroscopy (XPS). The results reported herein provide new insights into the design and synthesis of advanced electrode materials with various degrees
of structural defects for use in high-performance energy storage and conversion devices.

their complex structure consists of two different local structures


1. Introduction of monoclinic (space group of C/2m) Li2MnO3 and rhombohe-
Advanced electrode materials that deliver more energy at rapid dral (space group of R3 m) LiMO2 at an atomic scale as deter-
charge and discharge rates are essential for further enhance- mined by X-ray diffraction and TEM analysis.7–10 As a result, the
ment of the energy and power density of rechargeable lithium- Li2MnO3 component plays a crucial role in the electrochemistry
ion batteries in order to meet ongoing market demand for high- of xLi2MnO3$(1  x)LiMO2 electrodes, which not only improves
performance lithium-ion batteries in the future. Recently, the the structural stability of layered LiMO2 at high potentials
layered lithium-rich transition-metal oxides with the general resulting in excellent cyclic performance, and high safety and
formula xLi2MnO3$(1  x)LiMO2 (M ¼ Mn, Co, Ni, etc.) have thermal stability, but also serves as the electrochemical active
been the focus of intense research interest as one of the most phase for lithium extraction when charged above 4.5 V vs. Li/Li+
promising cathode materials for high-energy-density lithium- giving rise to a characteristic voltage plateau at ca. 4.5 V on the
ion batteries due to their high reversible capacities of up to initial charge and hence high reversible capacities. Intensive
250 mA h g1 or more at low cost,1–7 which represents an investigations on the reaction mechanisms of the layered
important milestone in materials design for advanced lithium- lithium-rich transition-metal oxides have revealed that the high
ion batteries. capacity may originate either from the removal of oxygen from
However, the microstructure of such materials is quite the Li2MnO3 component accompanied by lithium extraction5,7,11
complex and is still under debate. Recent studies on layered or from the surface reaction through electrode/electrolyte
lithium-rich transition-metal oxides have demonstrated that reduction and/or Li+/H+ exchange.12–14
Although much progress has been achieved on the lithium-
rich transition-metal oxides, the relationship between the
Department of Chemistry and Shanghai Key Laboratory of Molecular Catalysis and
Innovative Materials, Institute of New Energy, Fudan University, Shanghai 200433,
microstructure e.g. stacking faults and electrochemical prop-
People's Republic of China. E-mail: yyxia@fudan.edu.cn; Fax: +86-21-51630318 erties of lithium-rich transition-metal oxides remains unclear.

This journal is © The Royal Society of Chemistry 2014 Energy Environ. Sci., 2014, 7, 705–714 | 705
View Article Online

Energy & Environmental Science Paper

Very recently, our work demonstrated that the 0.5Li2M- Materials characterization
nO3$0.5LiMn1/3Ni1/3Co1/3O2 nanomaterial prepared via a facile
Powder X-ray diffraction (XRD) measurements were performed
molten-salt strategy using a KCl ux delivered a high reversible
on a Rigaku D/MAX-IIA X-ray diffractometer equipped with a Cu-
capacity of 310 mA h g1 with signicant enhancement in initial
Ka radiation source (Bruker, Germany). And the XRD peak
coulombic efficiency (87%) at room temperature, exhibited
position was calibrated with graphite as a reference. Rietveld
superior rate capability and showed improved electrochemical
renement analysis was carried out using Topas soware
properties in particular at low temperatures.15 However, the
package (Total Pattern Analysis Solution Soware, version 4.2) in
structural properties behind the battery performance are not the 2q range of 10–100 at 2q step-scan intervals of 0.02 with a
well understood, whereas it may be helpful to note that a broad step time of 2 s. The particle morphologies of the samples were
and low-intensity superlattice peak is observed in the XRD
observed with a JEOL JSM-6390 scanning electron microscope
pattern, corresponding to the integrated monoclinic (C2/m)
(SEM) operating at 25 kV. Transmission electron microscopy
Li2MnO3-like component with stacking faults. In addition,
(TEM) images and selected area electron diffraction (SAED)
structural defects have been suggested to play a key role in the
patterns were examined with a JEOL JEM-2100 electron micro-
electrochemistry of the Li2MnO3 phase. Prior studies on the
scope equipped with a double tilt specimen holder. Specimens
Published on 11 November 2013. Downloaded on 1/4/2023 6:50:19 PM.

pure Li2MnO3 have shown that Li2MnO3 nanomaterials with


for observations were powder samples supported on a copper
various degrees of structural defects deliver high reversible micro-grid. Raman measurements were performed in a back-
capacities of up to 200 mA h g1, which is probably attributed to scattered conguration using a micro-Raman spectrometer
the material's defect chemistry.16 Recently, work by Yu et al.
LabRam HR 800 equipped with an Ar+ laser (l ¼ 632.8 nm) from
proposed that larger charge–discharge capacity could be
Jobin-Yvon Horiba Inc. at room temperature. X-ray photoelec-
obtained from pure Li2MnO3 materials with smaller particle
tron spectroscopy (XPS) measurements were performed on a PHI
size and a larger amount of stacking faults.17 Therefore, it is of
5000C ESCA system with a Mg Ka source operating at 14.0 kV and
great importance, but a challenge, to get a better insight into the
25 mA. The energies of the spectra were calibrated by the binding
relationship between the microstructure e.g. stacking faults and
energy of C 1s at 284.6 eV as a reference. The specic surface area
the electrochemical performance of the lithium-rich transition- of the as-prepared materials was characterized by the Brunauer–
metal oxides. In this work, we extensively studied the electro- Emmett–Teller (BET) method (Quadrasorb SI Automated Surface
chemical performance of Li1.2Mn0.54Ni0.13Co0.13O2 electrodes
Area and Pore Size Analyzer, multipoint mode).
with various degrees of stacking faults, which were prepared via
a facile molten-salt strategy using a variety of uxes including
Electrochemical tests
KCl, Li2CO3, and LiNO3.
The electrochemical measurements were performed in CR2016-
type coin cells. For the preparation of the Li1.2Mn0.54Ni0.13Co0.13O2
2. Experimental electrodes, 80 wt% of active material, 10 wt% carbon black
Materials synthesis conductive materials, and 10 wt% polyvinylidene diuoride (PVDF)
The Li1.2Mn0.54Ni0.13Co0.13O2 materials were prepared in KCl binder were mixed in N-methyl-2-pyrrolidone (NMP) solvent. Then,
(melting point, 780  C), Li2CO3 (melting point, 630  C), or LiNO3 the slurry was cast onto an aluminum-foil current collector. Aer
(melting point, 260  C) uxes via the molten-salt method. For coating, the electrodes were dried at 80  C for 10 h to remove the
the synthesis in the KCl ux, the raw materials were stoichio- solvent before being pressed. The electrode lm was punched in
metric amounts of Li2CO3 (5% excess of lithium to compensate the form of disks, typically with a diameter of 12 mm, and then
for any evaporative lithium loss during the calcination) and the dried at 80  C for 12 h under vacuum. The typical mass loading of
Ni–Co–Mn oxide precursor (Ni : Co : Mn molar ratio of the active material of the working electrodes was approximately
0.13 : 0.13 : 0.54). For the syntheses in Li2CO3 and LiNO3 uxes, 5 mg. The cells were assembled with the cathode as prepared,
the raw material was the previously mentioned Ni–Co–Mn oxide lithium metal as the anode, and Celgard 2300 lm as the separator
precursor. The mole ratio between the eutectic mixture and the in a glove-box lled with pure argon. The electrolyte solution was
Ni–Co–Mn oxide precursor was xed at 4 : 1. Aer well mixing, 1 M LiPF6 dissolved in ethylene carbonate (EC)–dimethyl carbonate
the mixture was placed in an alumina crucible, heated to the (DMC)–ethyl methyl carbonate (EMC) (1 : 1 : 1 by volume). Galva-
desired temperature (ranged from 700 to 900  C) in air in a box nostatic charge–discharge experiments were performed between
furnace at a heating rate of 5  C min1 for 10 h, and was then 2.0 and 4.8 V at a current density of 20 mA g1 at room temperature
allowed to cool to room temperature naturally. Finally, the (30  C) with a LAND CT2001A Battery Cycler (Wuhan, China).
product was washed with distilled water several times and Lithium intercalation into the working electrode was referred to as
ltered, and the powder was dried in air in an oven at 120  C discharge, and the de-intercalation as charge. The cell capacity was
for 24 h. calculated based on the weight of the active material.
The Ni–Co–Mn precursor was prepared via a co-precipitation
method. A 2 M aqueous solution of NiSO4, CoSO4, and MnSO4 3. Results and discussion
(Ni : Co : Mn molar ratio of 0.13 : 0.13 : 0.54) was pumped into
a 2 M aqueous solution of Na2CO3 that contained 0.2 M NH4OH Synthesis of Li1.18Mn0.54Ni0.13Co0.13O2 in a KCl ux
at 50  C. Finally, the obtained precursor was ltered, dried, and Fig. 1 shows the powder X-ray diffraction (XRD) patterns of the
then sintered at 500  C for 5 h. Li1.18Mn0.54Ni0.13Co0.13O2 products prepared in a KCl ux at

706 | Energy Environ. Sci., 2014, 7, 705–714 This journal is © The Royal Society of Chemistry 2014
View Article Online

Paper Energy & Environmental Science


Published on 11 November 2013. Downloaded on 1/4/2023 6:50:19 PM.

Fig. 1 XRD patterns of the as-prepared Li1.18Mn0.54Ni0.13Co0.13O2


powders prepared at different temperatures for 10 h using a KCl flux/ Fig. 2 SEM and TEM images of the as-prepared Li1.18Mn0.54Ni0.13-
metal oxide precursor ratio of 4. The arrow marks impurity peaks Co0.13O2 powders prepared at different temperatures for 10 h using a
related to a spinel phase. The inset highlights the superlattice peaks in KCl flux/metal oxide precursor ratio of 4: (a) 700  C; (b) 800  C; (c)
the 2q range of 20–25 . 900  C; (d) 800  C.

various temperatures, ranging from 700  C to 900  C. The positions are indicated in Fig. 1 and are consistent with the
observed sharp diffraction peaks could be indexed to a hexag- results of previous studies.7,18,19 The extent of lithium-ion
onal unit cell with a trigonal (R3 m) space-group symmetry, ordering in the transition metal layers is reected by the peak
except for those broadened and low-intensity superlattice lines intensities of the (020) reections. As expected, the intensity of
in the 2-theta range of 20–25 . The superlattice peaks are the this peak increased signicantly as the heat-treatment temper-
features of the integrated monoclinic (C2/m) Li2MnO3-like atures were increased (Fig. 1, inset). However, at heat-treatment
component owing to lithium-cation ordering in the transition- temperatures greater than 900  C, spinel-phase impurities were
metal layer. All these characteristic peaks were present in the observed, as shown in Fig. 1c. These results indicate that 800  C
XRD patterns of all of the ux-synthesized products; their is a suitable temperature for achieving well-developed

Fig. 3 (a–c) Typical charge–discharge curves examined at a current density of 20 mA g1 for the Li1.18Mn0.54Ni0.13Co0.13O2 electrodes prepared
at different temperatures for 10 h with a KCl flux/metal oxide precursor ratio of 4 and (d) the cycling performance of samples prepared at 800  C.

This journal is © The Royal Society of Chemistry 2014 Energy Environ. Sci., 2014, 7, 705–714 | 707
View Article Online

Energy & Environmental Science Paper

pure-phase crystallization in a KCl ux via the molten-salt


method. The as-prepared pristine material obtained at 800  C
was characterized using inductively coupled plasma atomic
emission spectrometry (ICP-AES), which indicated that the
Li : Mn : Ni : Co ratio was 1.18 : 0.54 : 0.13 : 0.13.
The SEM images illustrated in Fig. 2a–c reveal that higher
heat-treatment temperatures led to larger particle sizes, sug-
gesting that the reaction temperature had a strong inuence on
the crystallite growth of Li1.18Mn0.54Ni0.13Co0.13O2 particles. As
shown in Fig. 2d, further analysis by transmission electron
microscopy (TEM) conrmed that the sample calcined at 800  C
consists of monodispersed irregular nanoparticles with diam-
eters of approximately 100–200 nm. The value of specic surface
area measured by the BET method was 12.4 m2 g1.
Published on 11 November 2013. Downloaded on 1/4/2023 6:50:19 PM.

Fig. 3a–c depict the typical charge–discharge curves of the as-


Fig. 4 XRD patterns of the as-prepared Li1.24Mn0.54Ni0.13Co0.13O2
prepared Li1.18Mn0.54Ni0.13Co0.13O2 electrodes at the 1st, 10th,
powders prepared at different temperatures for 10 h using a Li2CO3 flux/
and 30th cycles in the potential range of 2.0–4.8 V at room metal oxide precursor ratio of 4. * marks an unknown impurity peak. The
temperature. Upon initial charge to 4.8 V, the electrochemical inset highlights the superlattice peaks in the 2q range of 20–25 .
reaction of lithium with the Li1.18Mn0.54Ni0.13Co0.13O2 electrode
occurred in two distinct steps, as follows: the rst step
measured at voltage <4.4 V vs. Li/Li+ corresponds to lithium contrast to the sample prepared at 900  C in a KCl ux. The
extraction from the LiMn1/3Ni1/3Co1/3O2 component with the results of ICP-AES analyses indicate that the formula of the
concomitant oxidation of Ni2+ to Ni4+ and Co3+ to Co4+; and the product obtained at 800  C in a Li2CO3 ux was Li1.24Mn0.54-
second step commonly corresponds to the removal of lithium Ni0.13Co0.13O2, which is slightly Li-excess compared with its
from the Li2MnO3 component accompanied by oxygen evolu- nominal formula.
tion when the cell is charged above 4.4 V vs. Li/Li+.7,20,21 The Fig. 5 shows the morphology of the samples prepared at
electrode material prepared at 700  C showed poor electro- different temperatures in a Li2CO3 ux. The SEM images in
chemical performance; however, the sample synthesized at Fig. 5 show that the higher heat-treatment temperatures resul-
800  C in a KCl ux delivered a reversible discharge capacity ted in larger particle sizes. The TEM image in Fig. 5d conrms
of 310 mA h g1 on the rst cycle at a current density of that the sample prepared at 800  C in a Li2CO3 ux consists of
20 mA g1 at room temperature, which indicates a substantial monodispersed irregular particles with diameters on the order
improvement compared with the capacities reported in of 100–300 nm with the surface area of 16.3 m2 g1 measured by
previous studies. The typical charge–discharge curve of the the BET method.
electrode obtained at 900  C shows a unique voltage plateau at Fig. 6 demonstrates the typical electrochemical properties of
ca. 4.7 V, which arises from the spinel-phase impurities and is the as-prepared Li1.24Mn0.54Ni0.13Co0.13O2 electrodes synthe-
consistent with the XRD observations. The cyclic performance sized at different temperatures in a Li2CO3 ux when cycled
of the Li1.18Mn0.54Ni0.13Co0.13O2 electrode prepared in a KCl ux
at 800  C when cycled between 2.0 and 4.8 V at a current density
of 20 mA g1 is illustrated in Fig. 3d. The electrode retained 69%
of its initial capacity aer 100 cycles; this result implies that the
Li1.18Mn0.54Ni0.13Co0.13O2 electrode exhibits a poor cycle life
ascribed to the increasing side reactions between the electrode
and electrolyte interface area, which are, in turn, attributed to
the nanosized particles.22

Synthesis of Li1.24Mn0.54Ni0.13Co0.13O2 in a Li2CO3 ux


The powder X-ray diffraction (XRD) patterns of the as-prepared
products synthesized at different temperatures in a Li2CO3 ux
are shown in Fig. 4. The XRD patterns conrm that each pure
phase could be obtained using a Li2CO3 ux. Notably, the
observed weak diffraction peaks in the 2-theta range of 20–25
were sharper in comparison with those in the XRD patterns of
the samples prepared in a KCl ux, which suggests a greater
Fig. 5 SEM and TEM images of the as-prepared Li1.24Mn0.54Ni0.13-
extent of lithium-ion ordering in the transition-metal layers and Co0.13O2 powders prepared at different temperatures for 10 h using a
the characteristic Li2MnO3 component. In addition, the product Li2CO3 flux/metal oxide precursor ratio of 4: (a) 700  C; (b) 800  C; (c)
synthesized at 900  C in a Li2CO3 ux shows no spinel phase, in 900  C; (d) 800  C.

708 | Energy Environ. Sci., 2014, 7, 705–714 This journal is © The Royal Society of Chemistry 2014
View Article Online

Paper Energy & Environmental Science


Published on 11 November 2013. Downloaded on 1/4/2023 6:50:19 PM.

Fig. 6 (a–c) Typical charge–discharge curves examined at a current density of 20 mA g1 for the Li1.24Mn0.54Ni0.13Co0.13O2 electrodes prepared at
different temperatures for 10 h with a Li2CO3 flux/metal oxide precursor ratio of 4 and (d) the cycling performance of samples prepared at 800  C.

between 2.0 and 4.8 V at a current density of 20 mA g1. As


depicted in Fig. 6a–c, all the samples prepared in a Li2CO3 ux
showed a slight characteristic of a voltage plateau at ca. 4.5 V.
Aer several initial cycles for complete activation, the electrode
prepared at 800  C in a Li2CO3 ux delivered the highest
reversible capacity of 260 mA h g1, as shown in Fig. 6b. The
cycle life of the as-prepared Li1.24Mn0.54Ni0.13Co0.13O2 electrode
synthesized in a Li2CO3 ux at 800  C between 2.0 and 4.8 V at a
current density of 20 mA g1 is shown in Fig. 6d. Upon cycling,
the reversible capacity increased during the initial cycles, then
decayed aer 5 cycles due to electrochemical activation of the
Li2MnO3 component. The electrode retained 63% of its highest
capacity aer 100 cycles, indicating the poor cycling perfor-
mance of the Li1.24Mn0.54Ni0.13Co0.13O2 electrode prepared via
the molten-salt method using a Li2CO3 ux.
Fig. 7 XRD patterns of the as-prepared Li1.39Mn0.54Ni0.13Co0.13O2
powders prepared at different temperatures for 10 h using a LiNO3
Synthesis of Li1.39Mn0.54Ni0.13Co0.13O2 in a LiNO3 ux flux/metal oxide precursor ratio of 4. The inset highlights the super-
lattice peaks in the 2q range of 20–25 .
Fig. 7 presents the XRD patterns of the as-prepared samples
synthesized at different temperatures in a LiNO3 ux, varying
from 700  C, 800  C, and 900  C, respectively. Obviously, all the
peaks in the XRD patterns could be indexed on the basis of which is signicantly Li-rich compared with its nominal
layered lithium-rich transition-metal oxide structures, in good formula.
agreement with the results of the samples prepared in Li2CO3 Fig. 8 shows the morphology of the obtained samples
uxes. However, the peak intensities of the weak reections at prepared at different temperatures in a LiNO3 ux. As is evident
20–25 were much stronger than those in the XRD patterns of in the SEM images, each sample is composed of monodispersed
the samples prepared in KCl and Li2CO3 uxes, which indicates irregular particles, and the particle size increased with
more lithium-ion ordering in the transition-metal layers as increasing synthesis temperature. The particle size of the
discussed in the later section. The ICP-AES analysis of the powder prepared at 800  C in a LiNO3 ux was approximately
sample obtained at 800  C in a LiNO3 ux showed that the 100–400 nm in diameter. And the value of the specic surface
experimental Li : Mn : Ni : Co ratio is 1.39 : 0.54 : 0.13 : 0.13, area for the as-prepared material at 800  C was 14.9 m2 g1.

This journal is © The Royal Society of Chemistry 2014 Energy Environ. Sci., 2014, 7, 705–714 | 709
View Article Online

Energy & Environmental Science Paper

sufficiently electrochemically activated, the sample synthesized


at 800  C using a LiNO3 ux via the molten-salt method showed
the highest reversible capacity of 210 mA h g1 among the
materials prepared at three different synthesis temperatures.
Fig. 9d shows the cyclic performance of the Li1.39Mn0.54Ni0.13-
Co0.13O2 electrode prepared using a LiNO3 ux at 800  C in the
potential range between 2.0 and 4.8 V at a current density of
20 mA g1 at room temperature (30  C). Similar to the reversible
capacity of the electrode prepared at 800  C using a Li2CO3 ux,
the reversible capacity of the electrode synthesized using a
LiNO3 ux increased during the initial cycles, then decayed
aer 25 cycles, as shown in Fig. 9d. The electrochemical acti-
vation of the Li2MnO3 component required additional cycles
compared to the samples prepared at 800  C using KCl and
Li2CO3 uxes. The electrode prepared at 800  C using a LiNO3
Published on 11 November 2013. Downloaded on 1/4/2023 6:50:19 PM.

Fig. 8 SEM and TEM images of the as-prepared Li1.39Mn0.54Ni0.13- ux retained 64% of its highest reversible initial capacity aer
Co0.13O2 powders prepared at different temperatures for 10 h using a 100 cycles.
LiNO3 flux/metal oxide precursor ratio of 4: (a) 700  C; (b) 800  C; (c)
900  C; (d) 800  C.
General discussion
Based on the previously discussed results, the samples prepared
Fig. 9a–c illustrate the typical 1st, 10th, and 30th charge– in KCl, Li2CO3 and LiNO3 uxes show remarkably different
discharge proles of Li1.39Mn0.54Ni0.13Co0.13O2 electrodes charge capacity at the characteristic voltage plateau during the
prepared at different temperatures in a LiNO3 ux between 2.0 rst cycle, implying the different extent of electrochemical
and 4.8 V at a current density of 20 mA g1 at room temperature. activation of the Li2MnO3 component when charged to 4.5 V or
Obviously, the charge–discharge curves of all the electrodes more during the initial cycle. The electrochemical properties of
prepared using a LiNO3 ux exhibited an insignicant voltage the as-prepared samples are strongly inuenced by the micro-
plateau at 4.5 V during the initial cycles, which is remarkably structure of the Li1.2Mn0.54Ni0.13Co0.13O2 materials, which is
different from the charge–discharge proles of the electrodes associated with the molten-salt type. To shed some light on this
prepared using KCl and Li2CO3 uxes. Aer the electrode was issue, further structural characterization of the samples

Fig. 9 (a–c) Typical charge–discharge curves examined at a current density of 20 mA g1 for the Li1.39Mn0.54Ni0.13Co0.13O2 electrodes prepared
at different temperatures for 10 h with a LiNO3 flux/metal oxide precursor ratio of 4 and (d) the cycling performance of samples prepared at
800  C.

710 | Energy Environ. Sci., 2014, 7, 705–714 This journal is © The Royal Society of Chemistry 2014
View Article Online

Paper Energy & Environmental Science


Published on 11 November 2013. Downloaded on 1/4/2023 6:50:19 PM.

Fig. 10 (a–c) Rietveld plots for the lithium-rich transition-metal oxides prepared at 800  C in different molten-salt fluxes: (a) KCl; (b) Li2CO3; (c)
LiNO3. The calculated patterns are shown by red solid curves; black dots show the observed intensities. The differences between the observed
and calculated intensities are presented by blue curves. And (d) XRD patterns in the 2q range of 20–30 of the lithium-rich transition-metal oxides
prepared at 800  C in different molten-salt fluxes.

prepared at 800  C in different molten-salt uxes was performed calculated fractions for the Li2MnO3 component agrees well
employing the Rietveld renement method with the Topas with the nominal composition of the prepared material, indi-
soware. The tting patterns are shown in Fig. 10a–c and the cating the approximate phase composition of the as-synthe-
lattice parameters calculated on the basis of the R3 m symmetry sized materials.
are listed in Table 1. The Rietveld renement using the R3 m and Raman spectroscopy is a vibrational technique that provides
C2/m symmetry provides reasonably good tting based on the structural properties at the atomic scale and has also been
Bragg R-factors, which is commonly used in the renements of recognized as one of the most sensitive tools to detect unique
these complex structurally integrated materials.23,24 The general molecular and crystalline information by determination of
trend of the c/a ratio provides information about the effect of frequencies of normal vibrations.27 Fig. 11 shows Raman
molten-salt type on the interlayer spacing of the closely packed spectra of the samples prepared at 800  C in KCl, Li2CO3, and
structure. The decrease in the value of the a parameter is LiNO3 uxes to check the nature and composition of the as-
ascribed to the increasing Li+ ion content in the octahedral sites prepared products as a function of the synthesis conditions. As
consistent with ICP-AES results and the increasing M4+ ion illustrated in Fig. 11, all the crystal structures give rise to the
content because tetravalent ions have a smaller ionic radius in same Raman ngerprint: that is, three Raman bands are
octahedral coordination (Mn4+: 0.54 Å, Ni4+: 0.48 Å, Co4+: 0.53 Å) observed at ca. 593, 474, and 425 cm1. This result is in good
compared to high-spin Mn3+ (0.65 Å), low-spin Ni3+ (0.56 Å), Ni2+ agreement with those reported previously23 and is consistent
(0.70 Å), high-spin Co3+ (0.61 Å), and low-spin Co2+ (0.65 Å), all with the theoretical prediction given for a hexagonal (R3 m) and
of which might be present at various levels in these highly a monoclinic (C2/m) crystal,27 respectively. In general, there are
complex integrated structures.25,26 Note that each of the two predicted Raman-active vibrational modes in the ideal
layered lithium transition metal oxide with R3 m symmetry: A1g
with the symmetrical stretching of M–O and Eg with the
Table 1 Refined lattice parameters of the lithium-rich transition-metal symmetrical deformation, which corresponds to two sharp
oxides prepared at 800  C in different molten-salt fluxes
Raman peaks near 593 cm1 and 474 cm1, respectively. An
Fluxes Fractionsa (wt%) a (Å) c (Å) c/a (Å) additional small Raman band at 425 cm1 originates from the
Li2MnO3-like structure due to the reduced local symmetry of
KCl 56.2% 2.8517(1) 14.2141(1) 4.9844(2) C2/m rather than R3 m.28 This result provided by Raman
Li2CO3 59.1% 2.8470(7) 14.2165(0) 4.9933(8) spectroscopy further conrms that the as-synthesized materials
LiNO3 59.7% 2.8435(8) 14.2057(9) 4.9957(4)
in KCl, Li2CO3, and LiNO3 uxes show approximate composi-
a
The calculated fractions for the Li2MnO3 component. tion and phase components with a nominal formula

This journal is © The Royal Society of Chemistry 2014 Energy Environ. Sci., 2014, 7, 705–714 | 711
View Article Online

Energy & Environmental Science Paper


Published on 11 November 2013. Downloaded on 1/4/2023 6:50:19 PM.

Fig. 11 Raman spectra of the Li1.2Mn0.54Ni0.13Co0.13O2 materials


prepared at 800  C in a: (a) KCl, (b) Li2CO3, and (c) LiNO3 flux.

Li1.2Mn0.54Ni0.13Co0.13O2. X-ray photoelectron spectroscopy


(XPS) has been utilized to examine the oxidation state of tran-
sition metals in the surface of the as-prepared materials. Fig. 12
shows the XPS spectra for Mn 2p, Ni 2p, and Co 2p regions
observed for samples prepared at 800  C in KCl, Li2CO3, and
LiNO3 uxes, respectively. As seen in Fig. 12, two peaks are
observed, which can be assigned to the manganese 2p3/2 at
642 eV and 2p1/2 at 654 eV and agree well with the literature data
on MnO2.14 However, no signicant difference is observed by
XPS for the as-prepared materials in different molten salt uxes
in this experiment. From the Ni 2p and Co 2p XPS spectra, the
observations are consistent with the results in the Mn 2p
spectra as shown in Fig. 12, indicating no difference in the
oxidation state of transition metals and surface states.
Therefore, based on the SEM observations, XRD analysis,
ICP-AES measurements, Raman and XPS analysis, it can be
proposed that the various electrochemical properties of the as-
prepared samples may be attributed to the amount of stacking
faults in the crystal structure of the prepared materials.
Recently, Yu et al. have demonstrated that larger charge–
discharge capacity is obtained from Li2MnO3 materials with
smaller particle size and a larger amount of stacking faults.17 As
expected, the increase in the amount of stacking faults could
result in the broadening of the superlattice peaks in the 2q
range of 20–25 .29 As shown in Fig. 10d, only a single broad peak Fig. 12 XPS spectra for Mn 2p, Co 2p, and Ni 2p observed for samples
at 20.9 is observed in the XRD pattern of the sample prepared prepared at 800  C in a: (a) KCl, (b) Li2CO3, and (c) LiNO3 flux,
respectively.
at 800  C in a KCl ux; however, the superlattice reections of
the sample prepared at 800  C in a LiNO3 ux were much more
sharp than those in the XRD patterns of the samples prepared
in KCl and Li2CO3 uxes, which indicates a decrease in the the second synthesized at 800  C for 10 h in a LiNO3 ux. For
amount of stacking faults and fewer structural defects in the simplicity, the typical diffraction spots were indexed in terms of
Li2MnO3 structure, thereby making it more difficult for elec- the R3m structure with hexagonal notation. The location of the
trochemical activation and delivering less charge–discharge diffuse streaks conrms the existence of a superstructure and
capacity. This point was further examined using the SAED thus lithium ordering in the transition-metal layer, as pointed
method. out by arrows. Note that no spots are distinguishable for the
Electron diffraction patterns collected along the [11 0] former material, whereas weak spots localized on the diffuse
hexagonal zone axis are presented in Fig. 13 for our two extreme streaks are clearly seen in the pattern of the latter one. Weill
samples: the one prepared at 800  C for 10 h in a KCl ux and et al. proposed that diffuse streaks in the electron diffraction

712 | Energy Environ. Sci., 2014, 7, 705–714 This journal is © The Royal Society of Chemistry 2014
View Article Online

Paper Energy & Environmental Science

of 310 mA h g1. Although a well-crystallized Li1.39Mn0.54-


Ni0.13Co0.13O2 sample with a smaller amount of stacking faults
and an excess of lithium could be prepared at 800  C using an
oxidizing LiNO3 ux, the electrode exhibited the worst electro-
chemical performance among the samples synthesized using
the different uxes and showed an insignicant voltage plateau
at 4.5 V when initially charged to 4.8 V. Moreover, the sample
obtained using a Li2CO3 ux at 800  C exhibited electro-
chemical behavior intermediate between those of the samples
prepared in a KCl ux and in a LiNO3 ux.
As a result, the state of the Li2MnO3 component plays a
crucial role in the electrochemistry of layered lithium-rich
transition-metal oxides as the electrochemical active phase,
which not only improves the structural stability of layered
Published on 11 November 2013. Downloaded on 1/4/2023 6:50:19 PM.

LiMO2 at high potentials resulting in high reversible capacities


and excellent cyclic performance, but also shows disadvantages
including irreversible capacity loss during the initial cycle, poor
rate capability, and voltage shape change during the electro-
chemical cycling. The results reported herein provide new
insights into the design and synthesis of high-performance
layered lithium-rich transition-metal oxides. The dependence of
the electrochemical properties of layered lithium-rich oxides on
their microstructure requires additional work to achieve a
better understanding of the behaviors of such materials.
Fig. 13 SAED patterns of the Li1.2Mn0.54Ni0.13Co0.13O2 samples
prepared at 800  C for 10 h in different molten-salt fluxes: (a) KCl; (b)
LiNO3. (c) Cyclic performance of the as-prepared samples at a current Acknowledgements
density of 20 mA g1.
This work was partially supported by the State Key Basic
Research Program of PRC (2011CB935903), the National
patterns could be the signatures of stacking faults in the Natural Science Foundation of China (20925312), and the
ordered cationic planes occurred along the c monoclinic axis. It Shanghai Science & Technology Committee (10JC1401500 and
is relevant that spots affected by the diffusion become the 08DZ2270500).
continuous diffuse streaks when the amount of stacking faults
along the c monoclinic axis increases.30 Therefore, the weak
spots localized on the diffuse streaks in the electron diffraction
References
pattern of the sample prepared at 800  C in a LiNO3 ux indicate 1 M. M. Thackeray, C. Wolverton and E. D. Isaacs, Energy
that a decrease in the amount of stacking faults of the ordered Environ. Sci., 2012, 5, 7854–7863.
cationic planes occurred along the c monoclinic axis, which is in 2 B. L. Ellis, K. T. Lee and L. F. Nazar, Chem. Mater., 2010, 22,
good agreement with the previous XRD results. A more detailed 691–714.
TEM study is currently under way. 3 T. Ohzuku, M. Nagayama, K. Tsuji and K. Ariyoshi, J. Mater.
Chem., 2011, 21, 10179–10188.
4. Conclusions 4 Z. Lu, D. D. MacNeil and J. R. Dahn, Electrochem. Solid-State
Lett., 2001, 4, A191–A194.
Layered lithium-rich transition-metal oxides with various 5 Z. H. Lu and J. R. Dahn, J. Electrochem. Soc., 2002, 149, A815–
degrees of stacking faults were successfully prepared via a A822.
molten-salt method using KCl, Li2CO3, and LiNO3 uxes. The 6 M. M. Thackeray, C. S. Johnson, J. T. Vaughey, N. Li and
frequency of the stacking faults is highly dependent on the S. A. Hackney, J. Mater. Chem., 2005, 15, 2257–2267.
temperature and molten salt type. During the reaction, the KCl 7 M. M. Thackeray, S. H. Kang, C. S. Johnson, J. T. Vaughey,
ux not only serves as a mineralizing agent to achieve high R. Benedek and S. A. Hackney, J. Mater. Chem., 2007, 17,
crystallinity, but also acts as an inactive liquid reaction medium 3112–3125.
that facilitates molecular-level mixing of the reactants. 8 J. S. Kim, C. S. Johnson, J. T. Vaughey, M. M. Thackeray and
However, the Li2CO3 and the oxidizing LiNO3 uxes further S. A. Hackney, Chem. Mater., 2004, 16, 1996–2006.
participate in the reaction by providing Li ions. In an inactive 9 J. Bareno, C. H. Lei, J. G. Wen, S. H. Kang, I. Petrov and
KCl ux, layered Li1.18Mn0.54Ni0.13Co0.13O2 nanoparticles with a D. P. Abraham, Adv. Mater., 2010, 22, 1122–1127.
larger amount of stacking faults and slight lithium deciency 10 H. Yu, R. Ishikawa, Y. G. So, N. Shibata, T. Kudo, H. Zhou
were obtained at 800  C and delivered a high reversible capacity and Y. Ikuhara, Angew. Chem., Int. Ed., 2013, 52, 1–6.

This journal is © The Royal Society of Chemistry 2014 Energy Environ. Sci., 2014, 7, 705–714 | 713
View Article Online

Energy & Environmental Science Paper

11 A. R. Armstrong, M. Holzapfel, P. Novak, C. S. Johnson, 21 M. G. Kim, M. Jo, Y. S. Hong and J. Cho, Chem. Commun.,
S. H. Kang, M. M. Thackeray and P. G. Bruce, J. Am. Chem. 2009, 218–220.
Soc., 2006, 128, 8694–8698. 22 Q. Y. Wang, J. Liu, A. V. Murugan and A. Manthiram, J.
12 M. Jiang, B. Key, Y. S. Meng and C. P. Grey, Chem. Mater., Mater. Chem., 2009, 19, 4965–4972.
2009, 21, 2733–2745. 23 F. Amalraj, D. Kovacheva, M. Talianker, L. Zeiri, J. Grinblat,
13 A. D. Robertson and P. G. Bruce, Chem. Mater., 2003, 15, N. Leifer, G. Goobes, B. Markovsky and D. Aurbach, J.
1984–1992. Electrochem. Soc., 2010, 157, A1121–A1130.
14 N. Yabuuchi, K. Yoshii, S. T. Myung, I. Nakai and S. Komaba, 24 A. Boulineau, L. Simonin, J. F. Colin, E. Canevet, L. Daniel
J. Am. Chem. Soc., 2011, 133, 4404–4419. and S. Patoux, Chem. Mater., 2012, 24, 3558–3566.
15 J. L. Liu, L. Chen, M. Hou, F. Wang, R. Che and Y. Y. Xia, J. 25 C. S. Johnson, N. Li, C. Leef, J. T. Vaughey and
Mater. Chem., 2012, 22, 25380–25387. M. M. Thackeray, Chem. Mater., 2008, 20, 6095–6106.
16 Y. S. Meng, G. Ceder, C. P. Grey, W. S. Yoon, M. Jiang, 26 R. D. Shannon and C. T. Prewitt, Acta Crystallogr., 1969, 25,
J. Breger and Y. Shao-Horn, Chem. Mater., 2005, 17, 2386– 925.
2394. 27 R. Baddour-Hadiean and J. Pereira-Ramos, Chem. Rev., 2010,
Published on 11 November 2013. Downloaded on 1/4/2023 6:50:19 PM.

17 D. Y. W. Yu, K. Yanagida, Y. Kato and H. Nakamura, J. 110, 1278–1319.


Electrochem. Soc., 2009, 156, A417–A424. 28 J. Hong, D. Seo, S. Kim, H. Gwon, S. T. Oh and K. Kang, J.
18 J. L. Liu, J. Wang and Y. Y. Xia, Electrochim. Acta, 2011, 56, Mater. Chem., 2010, 20, 10179–10186.
7392–7396. 29 J. Breger, M. Jiang, N. Dupre, Y. S. Meng, Y. Shao-Horn,
19 G. Z. Wei, X. Lu, F. S. Ke, L. Huang, J. T. Li, Z. X. Wang, G. Ceder and C. P. Grey, J. Solid State Chem., 2005, 178,
Z. Y. Zhou and S. G. Sun, Adv. Mater., 2010, 22, 4364–4367. 2575–2585.
20 S. H. Kang, J. Kim, M. E. Stoll, D. Abraham, Y. K. Sun and 30 A. Boulineau, L. Croguennec, C. Delmas and F. Weill, Solid
K. Amine, J. Power Sources, 2002, 112, 41–48. State Ionics, 2010, 180, 1652–1659.

714 | Energy Environ. Sci., 2014, 7, 705–714 This journal is © The Royal Society of Chemistry 2014

You might also like