F--substitution method by surface fluorination of LiMn2O4 has
been carried out by Yonezawa et al.16 In their report, LiMn2O4
has been treated with F2 gas at various temperatures and it has
been confirmed by X-ray photoelectron spectroscopy that the
surface of LiMn2O4 has been fluorinated. Under the optimal
fluorination condition, both discharge capacity and cycleability
of the surface fluorinated LiMn2O4 has been better than the
pristine LiMn2O4. The F--substitution in the bulk LiMn2O4 have
been carried out by Amatucci et al.17 They have synthesized
Li1+xMn2-xO4-d Fz and Li1+xAlyMn2-x-yO4-d Fz by adding LiF to
the starting materials and firing at a high temperature (800
◦C). The resulting spinel oxyfluoride exhibits higher discharge
capacity and better cycleability than LiMn2O4. However, it has
been reported that the amount of fluorine in spinel oxyfluoride is
influenced by the firing temperature and the amount of fluorine
is limited at high firing temperatures at which LiF can react.18
Hence, lowering the firing temperature should be important to
increase the fluorine content. Recently, Choi et al. reported the
synthesis of spinel oxyfluoride such as LiMn1.8Li0.1Ni0.1O3.927F0.073
at a low temperature (450 ◦C) by using spinel compounds and
NH4HF2 as the starting materials.19,20 In their reports, resulting
spinel oxyfluoride shows superior electrochemical performance
compared to LiMn2O4. These various reports are quite interesting,
since anion substitution shows the possibility to enhance the
performance of the spinel lithium manganese oxides. In order
to get the guide to improve the superior electrochemical prop-
erties, understanding the detailed local/electronic structures of
spinel oxyfluoride is essential. In contrast to the cation-doped
compounds, such detailed structures have not been clarified for
anion-substituted lithium manganese spinel oxides.
nitride (BN) to obtain optimum absorption jump (Dmt = 1), and the
mixture was pressed into pellets for the measurement. The spectra
were recorded by the transmission mode using Si(111) double
monochromators and a Rh-coated mirror for harmonic rejection
except that they were recorded in a high-vacuum chamber by
the total-electron-yield mode using a microchannel plate detector
for Mn L-, O K- and F K-edges. Fourier transformations were
performed using k3 weighting. The structural parameters were
determined by curve-fitting procedures using Rigaku REX2000
data analysis software.23 The effective backscattering amplitude
F
eff , phase correction g (k), and total central atom phase shift x
were calculated with the multiple-scattering theoretical calculation
program, FEFF8.20.24 The model of LiMn2O4 was selected to
input the starting parameters for the theoretical calculation. The
c(k) function was fitted according to the equation:
NS02 f (k ,p) exp(−2si2k 2 )exp(−2Ri / li )sin[2kRi +fi (k )]
c(k ) =
kRi2
∑
i
(1)
where N is the number of neighboring atoms, S02 is the amplitude
2
of c(k), R is the atomic distance to the neighboring atom, s is the
Debye–Waller (DW) factor, l is the mean free path, and f is the
total phase shift.
Charge-discharge measurements (1/24 C) were carried out
between 3.5 and 4.3 V vs. Li/Li+ at room temperature using
a three-electrode cell. Li foil was used as the counter and the
reference electrodes, and 1 mol dm-3 LiClO4/propylene carbonate
(PC) was used as an electrolyte solution. A working electrode was
a mixture of 75 wt% active material, 20 wt% acetylene black, and 5
wt% polyvinylidene difluoride (PVdF) coated onto an aluminum
current collector.
Delithiated powders for Ex-situ XANES and EXAFS mea-
surements were prepared by electrochemical method by using
the three-electrode cell mentioned above, except that the mixture
of 70 wt% active material, 20 wt% acetylene black, and 10 wt%
poly(tetrafluoroethylene) (PTFE), pressed into paste and wrapped
with Ni mesh, was used as the working electrode. Electrochemical
extraction of lithium ion was carried out with constant current
discharge (1/12 C). Ni mesh was removed from the paste for
preparing the delithated sample. For Mn and Ni K-edge, the
appropriate amount of each sample was mixed with BN to obtain
optimum absorption jump (Dmt = 1), and the mixture was pressed
into pellets for the measurements.
In the present paper, the local and electronic structures of the
partially F--substituted cation-doped lithium manganese spinel
oxides have been respectively investigated by X-ray absorption
near-edge structure (XANES) and extended X-ray absorption fine
structure (EXAFS) techniques to clarify the relationship between
the detailed structures and the discharge capacity profiles.
2. Experimental procedure
LiMn2-xLixO4 (x = 0, 0.05, 0.10, 0.15) and LiMn1.9-yLi0.1NiyO4
(y = 0, 0.05, 0.10, 0.15, 0.20) were synthesized by the solid-state
reactions. Li2CO3 (99.99% Kojundo Chemical Lab Industries),
NiO (99.9% Furuuchi Chemical Industries), and Mn2O3 obtained
by pre-heating of MnCO3 (99.99% Kojundo Chemical Lab
Industries) at 600 ◦C for 48 h were used as starting materials.
Required amounts of starting materials were mixed and heated at
800 ◦C for 2 days in air, and then cooled with a rate of 0.5 ◦C/min.
For partial fluorine substitution, the obtained LiMn1.9Li0.1Ni0.1O4
and NH4HF2 were mixed, and then fired at 450 ◦C for 5 h in air.20
Powder X-ray diffraction measurement (RINT-2200V) with Cu-
Ka radiation equipped with a graphite monochromator was used
to examine the crystal structure of resulting materials. All the
XRD analysis operated in the Bragg-Brentano geometry mode.
XANES and EXAFS measurements were performed to investigate
electronic and local structures of resulting materials on BL-
7C21 and BL-11A22 at Photon Factory, High Energy Accelerator
Research Organization Tsukuba, Japan.
3. Results and Discussion
3.1. Phase identification
The XRD patterns of LiMn2-xLixO4 (x = 0, 0.05, 0.10, 0.15)
and LiMn1.9-yLi0.1NiyO4 (y = 0, 0.05, 0.10, 0.15, 0.20) are shown
in Fig. 1. All peaks belonged to the spinel structure (Fd-3m)
and there were no peak of impurity. The lattice parameter was
calculated from the peak top method. Fig. 2 showed the plots of
the lattice parameter against the content of the low valent cation
(Li+ or Ni2+). In this figure, the lattice parameter was decreased
monotonically with increase of low valent cation content. Taking
the order of the ionic radius of Li+ > Ni2+ > Mn3+ (according
to Shannon’s ionic radii25) into account, it might happen that the
lattice parameter was increased with the increment of the low
For Mn and Ni K-edge, the appropriate amount of each sample
for XANES and EXAFS measurements was mixed with boron
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 9752–9764 | 9753
©