Degradation Analysis of LiCoO2 Positive Electrode Material of a Li-Ion Battery Using the Li K-Edge Signal Obtained from STEM-EELS Measurements

W1171771791 · 수집 2026-10-04 18:12:13

T12039 Electron and X-Ray Spectroscopy Techniques

CoT 라벨링

라벨링 전

PDF 파일

자동 다운로드한 PDF PDF 보기

기본 정보

DOI 제목 발행년도 저널 URL 페이지 열기 PDF URL PDF 열기 발행기관 피인용수 라이선스

추가 정보 초록 본문

Degradation Analysis of LiCoO2 Positive Electrode Material of a Li-Ion Battery Using

the Li K-Edge Signal Obtained from STEM-EELS Measurements _∗_

Noboru Taguchi, _†_ Hikari Sakaebe, Kuniaki Tatsumi, and Tomoki Akita

_Research_ _Institute_ _for_ _Ubiquitous_ _Energy_ _Devices,_
_National_ _Institute_ _of_ _Advanced_ _Industrial_ _Science_ _and_ _Technology_ _(AIST),_ _Ikeda,_ _Osaka_ _563-8577,_ _Japan_

(Received 9 January 2015; Accepted 16 April 2015; Published 13 June 2015)

The Li distribution in positive electrode material (LiCoO2) was observed after a charge-discharge cycling test by
analytical transmission electron microscopy (TEM). The electron energy-loss spectroscopy (EELS) imaging method
revealed an inhomogeneous Li distribution in the LiCoO2 particles after the charge-discharge cycling test. Using
the alternating least-squares (ALS) method for the data matrix of the EELS spectra, we obtained LiCoO2 and
Li-poor LiCoO2 spectra from the LiCoO2 particles. Around the degraded LiCoO2 particles, Co3O4 and fluoride
components were detected. It is suggested that inhomogeneous electrochemical reactions are one cause of LiCoO2
and electrolyte degradation.

[DOI: 10.1380/ejssnt.2015.284]

Keywords: Scanning transmission electron microscopy (STEM); Electron energy-loss spectroscopy (EELS); Cobalt oxides;
Li-ion battery; Li K-edge

I. INTRODUCTION

To study the degradation mechanism of Li-ion battery
materials, it is important to understand the atomic structures and chemical states of these materials. Analytical transmission electron microscopy (TEM) is useful for
studying these materials because of its high spatial resolution. In particular, direct analysis of Li ions using electron energy-loss spectroscopy (EELS) can be a powerful
method of investigating the chemical state of Li in Li-ion
battery materials [1–3]. However, M-edge signals from
Mn, Co, and Ni appear near the energy region of the Li
K-edge signal in EELS images, and the overlap between
the Li K-edge and M-edges of these transition metals often reduces the accuracy of the quantitative analysis or
interferes with the chemical analysis of the spectrum. The
aforementioned transition metals are widely used as electrodes in Li-ion batteries for charge compensation during
the Li insertion and extraction reactions. High-energyresolution EELS is effective in reducing the overlapping
problem in Li K-edge analysis.

By using EELS with a monochromator, the Li-ion distributions inside LiCoO2 particles in positive electrode
material with and without surface modification during the
charge-discharge cycle have been reported [4, 5]. However,
the products generated from the electrochemical reactions
in a LiCoO2 system have not been previously examined
by analytical TEM.

In the present study, Li K-edge analysis using
monochromated STEM-EELS was performed. In addition, maps of the Li intensity and of the reaction products
were also reconstructed from EELS images [4, 6]. To obtain chemical information about the sample, analyses not
only of the intensities of the EELS distributions but also
of their shapes were conducted as post-analysis. Such a

_∗_ This paper was presented at the 7th International Symposium

on Surface Science, Shimane Prefectural Convention Center (Kunibiki Messe), Matsue, Japan, November 2-6, 2014.

_†_ Corresponding author: n-taguchi@aist.go.jp

FIG. 1. (a) SEM image of LiCoO2 particles. (b) ADF-STEM
image of LiCoO2. Incident beam direction is _⟨_ 100 _⟩_ .

FIG. 2. Cycling test results for cells containing LiCoO2.

large matrix of spectral data is suitable for analysis via a
chemometrics approach. To obtain individual pure spectra from the series of spectra without a reference spectrum, we applied a self-modeling curve-resolution technique (the alternating least-squares (ALS) method) [7–11]
to the series of spectra to perform spectral decomposition.

(a) (b)

500 nm 500 nm

5 nm

ISSN 1348 0391 _⃝_ c 2015 The Surface Science Society of Japan (http://www sssj org/ejssnt) 284

Li K

energy-loss (eV)

70 nm

0.0

1.0

FIG. 3. Li-K/Co-M intensity ratio map for LiCoO2 after cycling test. Image contrast is normalized to Li K-edge/Co M-edge
intensity ratio of as-prepared LiCoO2. Intensity scale corresponds to _x_ in Li _x_ CoO2. Raw spectra of Li-K edge are shown for all
spots (A-D).

The degradation process of LiCoO2 was analyzed based
on the series of decomposed spectra.

II. EXPERIMENTAL

Small LiCoO2 particles, which are suitable for TEM
and EELS analyses, were synthesized by the Pechini process [12]. Co(CH3COO)2 _·_ 4H2O, LiCH3COO _·_ H2O, and
citric acid were mixed at 165 _◦_ C in an ethylene glycol
solution (Co/Li = 1:1 atomic ratio). The obtained gel
was pre-calcined at 400 _◦_ C in air to decompose the precursor. The precursor was calcined again at 750 _◦_ C for 1
h in air. The LiCoO2 particles were about 200 nm in size.

A composite containing 84 wt% active material
(LiCoO2), 8 wt% acetylene black (AB), and 8 wt%
polyvinylidene fluoride (PVDF) was used as the positive electrode. A sheet-type two-electrode cell with
16 mm _ϕ_ electrodes sealed in Al-laminated type film
was assembled using Li metal as the counter electrode and 1.0 mol _·_ dm _−_ 3 LiPF6/EC (ethylene carbonate+DMC(dimethyl carbonate) (1:1 in volume) as the
electrolyte. The cell was cycled in the voltage range between 3.0 V and 4.2 V with a 0.1 C charge-discharge rate
(1 C = 150 mA/g). The temperature around the cell in
the charge-discharge tests was 20 _◦_ C.

For the TEM observations, LiCoO2 powder was directly dispersed on a carbon film supported by a Cu
mesh. A TITAN 3 G2 60–300 electron microscope (FEI
Company) equipped with a Super X (Bruker Corporation) for energy-dispersive X-ray spectroscopy (EDS) and
a GIF Quantum ERS (Gatan Inc.) for EELS was operated at an accelerating voltage of 200 kV. The EDS and
EELS measurements were performed on LiCoO2 after the
10 _th_ discharge by a spectrum-imaging scheme based on
STEM observations. The process of spectrum imaging
refers to the acquisition of three-dimensional data that
contain both spectral and spatial information about the
specimen. Each EELS image integrates the signal from
the whole region along the beam direction at each pixel.
In the EELS acquisition for Li-K mapping, a monochromator was applied to achieve a higher energy resolution.
The FWHM of the zero-loss peak was less than 0.3 eV

with an energy dispersion of 0.05 eV/channel.

The ALS method was applied to these EELS data [7–
11, 13–15]. As an initial ALS estimation, the orthogonal
projection approach (OPA) was used [16, 17].

III. RESULTS AND DISCUSSION

Figure 1 (a) shows the SEM image of the synthesized
LiCoO2 particles. The primary particles display a partial
agglomerate formation, and the size of primary particles
is about 200 nm. Thus, smaller-sized particles, which
are suitable for TEM observations, were obtained. The
typical layered structure of LiCoO2 was also confirmed
by using X-ray diffraction (XRD) (not shown) and ADFSTEM observations (Fig. 1 (b)).

Figure 2 shows the cycling performance of the cell.
The discharge capacity of LiCoO2 depends on the particle size [18]. Much larger particles ( _∼_ 300 nm) generally
show much better electrochemical properties [4]. In the
present study, we treated the smaller particles as the typical model sample to represent the capacity fading and
irreducible capacity in the TEM analyses.

The Li K-edge EELS plots of the LiCoO2 particles after 10 _th_ discharge are shown in Fig. 3 (A-D) along with
the Li distribution. We obtained the spectra around the
Li K-edge in each pixel of the ADF-STEM image. This
information enabled the spatial distribution of Li to be
reconstructed from the Li K-edge spectra [4, 5]. We reconstructed the Li/Co intensity-ratio map using the series
of spectra in Fig. 3. Inhomogeneous Li distributions in the
particles were clearly observed. At spot D, the shape of
the spectra is the same as that of bare LiCoO2 [4, 19].
Compared with the intensity of the Li K-edge at spot D,
the corresponding intensity at spot A is slightly decreased.
Considerably decreased Li K-edge intensities are observed
at spots B and C, and the regions around these spots
are believed to have been damaged during the chargedischarge cycling.

In our previous study, the homogeneity of the Li-ion
distribution in Al- and Si-oxide-coated LiCoO2 agreed
well with the capacity retention after the cycling test [4].
Specifically, LiCoO2 with poor cycling had a decidedly

http://www sssj org/ejssnt (J Stage: http://www jstage jst go jp/browse/ejssnt/) 285

# Co O

70 nm 70 nm

# C

70 nm

FIG. 4. Co, O, and C EDS mapping results of LiCoO2 after
cycling test.

inhomogeneous Li distribution, while Al-oxide-coated
LiCoO2 with good cycling exhibited a quite homogeneous
Li distribution. The present LiCoO2 with poor cycling is
consistent with this trend, showing an inhomogeneous Li
distribution.

The results of the EDS mapping of the LiCoO2 particles
after the 10th discharge are shown in Fig. 4. The gross
intensity maps in the energy windows of Co, O, and C
are indicated. No intensity changes related to the Li-poor
regions are observed in the maps of either Co or O. Only

comp. 1 comp. 2

comp. 3 comp. 4

(a) comp. 1 (b) comp. 2

LiCoO2 poor Li region (deform)

(capacity loss)

60 65 70 75 80
energy-loss (eV)

(e) comp. 5

60 65 70 75 80
energy-loss (eV)

(d) comp. 4

60 65 70 75 80
energy-loss (eV)

FIG. 6. EELS distributions obtained from ALS analysis of
LiCoO2 after cycling test.

(a) Li extraction

Li + Li + Li + Li +

locally overcharge

surface

LiCoO2

(b) Li insertion

Li +
Li +

Li +

decomposition product
(electrolyte LiFx, LiOxFy )

Co3O4

Li +

Li + Li +

surface

comp. 5

BF-STEM

70 nm

FIG. 5. Component maps obtained from ALS. BF-STEM image is also shown.

FIG. 7. Schematic image of sample degradation process during
cycling test.

intensity changes due to variations in sample thickness
can be seen. Thus, the amounts of Co and O do not
change drastically during charge-discharge cycling.

In the present sample, the degraded LiCoO2 area can
be clearly seen from the Li K-edge EELS and Li distribution images. The information about the chemical state
near the damaged region is quite important in order to
investigate the reaction products. In addition to the Li
intensity analysis shown in Fig. 3, the fine structures of
the EELS distributions also yield chemical information
about the sample. On these data, we performed ALS
analysis in order to identify the primary spectral compo

286 http://www sssj org/ejssnt (J Stage: http://www jstage jst go jp/browse/ejssnt/)

nents, especially in the degraded regions. From the ALS
analysis, we identified five types of components from the
spectral images, and their intensity maps are shown in
Fig. 5. The intensity map of comp. 1 corresponds closely
to a Li-poor region. In contrast, comp. 2 corresponds to
a bare LiCoO2 region. In comps. 3 and 5, very similar regions are observed, as these components were mainly obtained around the same particle edge. Similarly, comp. 4
was also obtained from the region surrounding the particle.

The decomposed spectra for all of these components are
shown in Fig. 6. In the spectrum of comp. 1 (Fig. 6 (a)),
a weak Li K-edge peak can be seen around 62 eV. The Li
K-edge spectrum of delithiated LiCoO2 (Li0 _._ 5CoO2) corresponds closely to the shape of the comp. 1 spectrum [19].
The Co and O intensities did not drastically change from
the EDS results after the 10th discharge, which are shown
in Fig. 4. This result indicates that the Li-poor region
does not transform pure Co-oxide completely as a result
of the Li absence but rather creates Li-poor states, such
as a charged state in which Li is partially extracted. In
the BF-STEM image, the region with greater particle contrast corresponds closely to the region in which comp. 1 is
mainly observed. In general, BF-STEM images are sensitive to diffraction contrast, such as that introduced by
local strain or dislocation in the crystal. In LiCoO2, significant Li extraction causes substantial change in the lattice constant. Local states are considered to induce more
particle damage than are charged states. Thus, local area
deformation is believed to prevent normal diffusion of Li
ions in materials.

In comp. 3, the Li K-edge peak completely disappears, as shown in Fig. 6 (c), in which there is no peak
near 62 eV. This spectrum corresponds closely to that
of Co3O4 [19]. Based on the present EELS results, the
main product around the LiCoO2 particle is believed to be
Co3O4. The spectrum of comp. 2 in Fig. 6 (b) corresponds
to that of LiCoO2 and includes the sharp Li K-edge peak.
The intensity distribution of comp. 4 is that of the particles surrounding the LiCoO2, and the same distribution is
obtained from EDS analysis of carbon. Thus, comp. 4 is
thought to be a background signal from carbon from AB
(conductive additive), electrolytes, or contamination from
TEM sample preparation, among other sources (Fig. 6
(d)). Finally, the peak of comp. 5 in Fig. 6 (e) agrees
well with the LiF _x_ and LiO _x_ F _y_ peaks previously reported
by Muto _et_ _al_ . [10]. These F compounds are believed to
be decomposition products from the electrolyte LiPF6.
The reaction products around the particle surface were

[1] J. Graetz, C. C. Ahn, R. Yazami, and B. Fultz, J. Phys.

Chem. B 107, 2887 (2003).

[2] J. Kikkawa, T. Akita, M. Tabuchi, M. Shikano, K. Tat
sumi, and M. Kohyama, Electrochem. Solid-State Lett.
11, A183 (2008).

[3] F. Wang, J. Graetz, M. S. Moreno, C. Ma, L. Wu, V.

Volkov, and Y. Zhu, ACS Nano 5, 1190 (2011).

[4] N. Taguchi, T. Akita, H. Sakaebe, K. Tatsumi, and Z.

Ogumi, J. Electrochem. Soc. 160, A2293 (2013).

[5] N. Taguchi, H. Sakaebe, T. Akita, K. Tatsumi, and Z.

Ogumi, J. Electrochem. Soc. 161, A1521 (2014).

determined from the results of TEM-EELS analysis of a
LiNi0 _._ 8Co0 _._ 15Al0 _._ 05O2 sample, which was fabricated by a
focused ion beam (FIB). Results similar to those observed
in the degradation analysis of LiNi0 _._ 8Co0 _._ 15Al0 _._ 05O2 using STEM-EELS [8–11] were obtained even in the case of
LiCoO2.

Considering these ALS results, we can draw a schematic
image of the degradation process around a LiCoO2 particle, as is shown in Fig. 7. First, LiCoO2 (positive electrode) is delithiated (charge process), and Li is diffused
to Li metal (anode). In this delithiation process, excess
Li extraction may occur locally. Crystal deformation is
induced in the local area, and thus further Li diffusion is
prevented. Then, side reactions occur around the surface
of the damaged region, and Co3O4 particles or other electrolyte decomposition products are generated (Fig. 7 (a)).
In the discharge (Li insertion) process, Li returns to the
positive electrode. However, such reaction products and
deformed areas are considered unfavorable for the diffusion of Li ions. Finally, a Li-poor region (comp. 1) appears
in the LiCoO2 particle (Fig. 7 (b)).

IV. CONCLUSION

STEM-EELS analysis of the Li K-edge in LiCoO2 particles was conducted after charge-discharge cycling by using a monochromator. From the resulting spectra, the
LiCoO2 particles were confirmed to have inhomogeneous
Li distributions. ALS analysis revealed that the Li-poor
region was not Li-free Co oxide and that Co3O4 and electrolyte decomposition products were distributed across
the surface. Local overcharge states induced side reactions and particle deformation, preventing Li diffusion in
the Li insertion and exertion processes during the chargedischarge cycling. STEM-EELS imaging with ALS analysis was successfully applied in the degradation analysis
of LiCoO2 positive electrode material.

ACKNOWLEDGMENTS

Authors are grateful to Dr. M. Kitta at AIST for his
valuable comments and fruitful discussion. This work was
supported by the “Research and Development Initiative
for Scientific Innovation of New Generation Battery (RISING project)” of the New Energy and Industrial Technology Development Organization (NEDO, Japan).

[6] M. Colliex, J. Electron Microsc 45, 44 (1996).

[7] R. Tauler, Chemom. Intell. Lab. Systems 30, 133 (1995).

[8] S. Muto, T. Yoshida, and K. Tatsumi, Mater. Trans. 50,

964 (2009).

[9] S. Muto, Y. Sasano, K. Tatsumi, T. Sasaki, K. Horibuchi,

Y. Takeuchi, and Y. Ukyo, J. Electrochem. Soc. 156,
A371 (2009).

[10] S. Muto, K. Tatsumi, T. Sassaki, H. Kondo, T. Ohsuna,

K. Horibuchi, and Y. Takeuchi, Electrochem. Sol-State
Lett. 13, A115 (2010).

http://www sssj org/ejssnt (J Stage: http://www jstage jst go jp/browse/ejssnt/) 287

[11] S. Muto, K. Tatsumi, Y. Kojima, H. Oka, H. Kondo,

K. Horibuchi, and Y. Ukyo, J. Power Sources 205, 449
(2012).

[12] Z. S. Peng, C. R. Wan, and C. Y. Jiang, J. Power Sources

72, 215 (1998).

[13] R Development Core Team (2012): http://www.Rproject.org/.

[14] K. M. Mullen, http://CRAN.R-project.org/package=als

(2012).

[15] R. Furrer, D. Nychka, and S. Satin, http://CRAN.R
project.org/package=fields (2012).

[16] F. Cuesta Sanchez, J. Toft, B. Bogaert, and D. L. Massart,

Anal. Chem. 68, 79 (1996).

[17] R. Wehrens, _Chemometrics_ _with_ _R_ (Springer Berlin Hei
delberg, 2012).

[18] C.-H. Lu, H.-H. Chang, and Y.-K. Lin, Ceramics Interna
tional 30, 1641 (2004).

[19] J. Kikkawa, S. Terada, A. Gunji, M. Haruta, T. Nagai,

K. Kurashima, and K. Kimoto, Appl. Phys. Lett. 104,
114105 (2014).

288 http://www sssj org/ejssnt (J Stage: http://www jstage jst go jp/browse/ejssnt/)

본문 출처 PDF: PDF 보기 · 다른 PDF는 위 [PDF 업로드]로 올릴 수 있습니다.

활용가능성

이력 0건

아직 편집 이력이 없습니다. 저장할 때마다 이전 내용이 여기에 백업됩니다.

Degradation Analysis of LiCoO&lt;sub&gt;2 &lt;/sub&gt;Positive Electrode Material of a Li-Ion Battery Using the Li K-Edge Signal Obtained from STEM-EELS Measurements

CoT 라벨링

PDF 파일

이력 0건

Degradation Analysis of LiCoO<sub>2 </sub>Positive Electrode Material of a Li-Ion Battery Using the Li K-Edge Signal Obtained from STEM-EELS Measurements