Available online at [Link].
com
Solid State Ionics 179 (2008) 1794 – 1799
[Link]/locate/ssi
The effects of TiO2 coating on the electrochemical performance of
Li[Li0.2Mn0.54Ni0.13Co0.13]O2 cathode material for lithium-ion battery
J.M. Zheng, J. Li, Z.R. Zhang, X.J. Guo, Y. Yang ⁎
State Key Laboratory for Physical Chemistry of Solid Surfaces, Department of Chemistry, Xiamen University, Xiamen 361005, PR China
Received 10 July 2007; received in revised form 15 January 2008; accepted 20 January 2008
Abstract
TiO2-coated Li[Li0.2Mn0.54Ni0.13Co0.13]O2 materials have been synthesized and investigated as cathode materials for lithium-ion batteries at
both 25 °C and elevated temperature (55 °C). The structure and morphology of the coated samples were characterized and compared. The XRD
results indicate that lattice parameters of the materials did not change distinctly after surface coating. The SEM images demonstrate that the
surface of Li[Li0.2Mn0.54Ni0.13Co0.13]O2 samples were covered with nano-sized TiO2 particles. Differential scanning calorimetry (DSC) analysis
results show that thermal stability of the materials was improved. It is also shown that the irreversible capacity loss of the materials was obviously
reduced and their capacity retention behaviour was improved after surface modification.
© 2008 Elsevier B.V. All rights reserved.
Keywords: Lithium-rich cathode materials; TiO2-coating layer; Lithium-ion battery; Surface modifications; Electrochemical performance
1. Introduction cathode in this series is Li[Li0.2Mn0.54Ni0.13Co0.13]O2, which
delivered an initial discharge capacity of about 260 mAh/g at
The mass application in portable electronic devices has C/20 rate over the potential range of 2.0 to 4.8 V [5]. However,
highlighted the importance of high energy density, light weight, the Li[Li0.2Mn0.54Ni0.13Co0.13]O2 cathode delivered only
and reliable lithium-ion batteries. To date, LiCoO2 has been about 175 mAh/g at 2 C rate and exhibited a large irreversible
widely used as one of the cathode material in commercial lithium- capacity loss (ICL) of 75 mAh/g. Accordingly, it has to over-
ion batteries owing to its good electrochemical performance and come the large irreversible capacity loss of the materials in
its ease of preparation, but it also suffers high cost, toxicity, and the 1st charge–discharge cycle and the shortcoming of inher-
safety problems. ently low conductivity of the material prior to its commercial
Cheaper, higher capacity, and safer layered-cathode ma- application.
terial has been one of the focusing points in the study of elec- Surface modification of the LiCoO2 with oxides such as
trode materials. Recently, Dahn's group attempted to stabilize Al2O3 [6], ZrO2 [7,8], SnO2 [9], MgO [10], and LiNi0.8Co0.2O2
the layered structure by using a solid solution between Li2MnO3 with ZrO2 [11], TiO2 [12,13], La2O3 [14] have been tes-
and LiMO2, such as Li[Ni x Li(1 − 2x) / 3Mn(2 − x) / 3]O2 for Ni2+ ions tified to improve the cyclability of the materials when
and Li[M xLi(1 − x) / 3Mn(2 − 2x) / 3]O2 for Co3+ and Cr3+ ions [1–4]. charged to the higher cut-off potential. Surface-modified layered-
Very interestingly, a Li/Li[Ni x Li(1 − 2x) / 3Mn(2 − x) / 3]O2 cell with Li[Li(1 − x) / 3Mn(2 − x) / 3Ni x / 3Co x / 3]O2 cathodes with Al2O3 has
x =1/3 delivered a reversible capacity of about 230 mAh/g. Besides, also been proved to be with lower irreversible capacity loss in
the series of Li[Li(1 − x) / 3Mn(2 − x) / 3Ni x / 3Co x / 3]O2 cathodes be- comparison with the uncoated ones [5]. In addition, Al(OH)3
longing to a solid solution between layered-Li[Li1/3Mn2/3]O2 coating on Li[Li0.2Ni0.2Mn0.6]O2 cathode has also been shown to
and Li[Mn1/3Ni1/3Co1/3]O2 become attractive. One promising decrease capacity decay at high rate and thermal stability com-
pared to the uncoated system [15].
⁎ Corresponding author. Here, we present some results about the surface modifi-
E-mail address: yyang@[Link] (Y. Yang). cation of Li[Li0.2Mn0.54Ni0.13Co0.13]O2 with TiO2, and the
0167-2738/$ - see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/[Link].2008.01.091
J.M. Zheng et al. / Solid State Ionics 179 (2008) 1794–1799 1795
effects of TiO2 coating on the initial irreversible capacity
loss, the electrochemical performance and thermal stability of
Li[Li0.2Mn0.54Ni0.13Co0.13]O2.
2. Experimental
The precursor of Li[Li0.2Mn0.54Ni0.13Co0.13]O2 compound
such as M(OH)2 (M = Ni, Mn, Co) was synthesized by a mixed-
hydroxide method [16]. Then the M(OH)2 and stoichiometric
amount of LiOH·H2O were pressed into pellets and pre-calcined
at 480 °C for 10 h, then ground and made into new pellets. The
target compound was finally obtained by quenching the new
pellets after firing at 900 °C for 3 h [15].
For preparation of samples coated with various molar ratio of
TiO2, pristine Li[Li0.2Mn0.54Ni0.13Co0.13]O2 powder was dis- Fig. 1. XRD patterns of the uncoated and TiO2-coated Li[Li0.2Mn0.54Ni0.13Co0.13]O2
persed in N-methyl-2-pyrrolidone (NMP), and then Ti(OBu)4 samples.
was dripped into the suspension. The suspension was ball-
milled and then exposed in air at 60 °C for 24 h to accelerate and samples. XRD pattern of blank TiO2 prepared from the same
ensure the complete hydrolysis of Ti(OBu)4. Finally, the pre- process of hydrolysis and heating treatment is also presented for
cursors from above were heated at 450 °C for 5 h to form the comparison. Except for the superlattice peaks between 20° and
TiO2-coated samples. 25°, all peaks of the XRD patterns can be indexed as an O3-type
X-ray diffraction (XRD) of these samples was performed on structure. Calculated lattice parameters from Rietveld refine-
a PANalytical X'Pert diffractometer with Cu Kα radiation ment of Li[Li0.2Mn0.54Ni0.13Co0.13]O2 are a = 2.8535 Å and
operated at 40 kV and 30 mA. Data were collected in the 2θ c = 14.2372 Å. Excepts one weak broad peak corresponding to
range of 10–90° at 4°/min. Scanning electron microscopy TiO2 at ca. 25.5° in 2θ is discerned for the samples coated with
(SEM) studies were performed on LEO1530 (Oxford compa- 5.0 and 9.0 mol% of TiO2, respectively, no extra reflections are
ny). The thermal stability of the Li[Li0.2Mn0.54Ni0.13Co0.13]O2 observed for other samples. Compared with the strong dif-
electrodes at delithiated state of 4.8 V was examined by means fraction peak at ca. 25.5° of bulk TiO2, the weak and even
of DSC with a Netzsch STA 409PC thermal analysis system absence of the characteristic peaks is possibly due to the small
from 50 to 350 °C at a heating rate of 5 °C/min. XPS analysis quantities of the coating layer. The right panel of Fig. 1 shows the
was conducted using a Physical Electronics Quantum 2000 (003) diffraction peaks of the uncoated and coated samples. It is
ESCA spectrometer (USA) with monochromatic Al Kα 1486.6 eV shown that the (003) peak locates almost at the same value in 2θ,
radiation operated at 23.2 W in a vacuum of b10− 8 Torr. The indicating that the lattice parameter of ‘c’ has no obvious change
electrodes for XPS analysis were recovered from the coin-type after coating.
cells after two cycles of charge/discharge concomitant with a
potentiostatic charge process at 4.3 V for 50 h. 3.2. SEM images of TiO2-coated Li[Li0.2Mn0.54Ni0.13Co0.13]O2
Electrochemical measurements were carried out using CR2025
coin-type cells. The cathodes were prepared by coating a mixture Fig. 2 shows the SEM images of various TiO2-coated
containing 80% active materials, 10% acetylene black, 10% Li[Li0.2Mn0.54Ni0.13Co0.13]O2 samples. Obviously, the surface of
PVDF binder on circular aluminium foils, followed by drying at the uncoated Li[Li0.2Mn0.54Ni0.13Co0.13]O2 particles is smooth.
120 °C for 1 h. Electrochemical cells were assembled with the After coating, the number of the nano-sized TiO2 particles
cathodes as-prepared, metallic lithium foil as counter electrode, attaching on the surface of the parent material increases from
Celgard 2300 as separator and 1 M LiPF6 dissolved in ethyl sparse for 1.0 mol% to distinctive for 3.0 mol% and finally to
carbonate (EC) and dimethyl carbonate (DMC) (1:1 in volume) as superfluous for 5.0 mol% and 9.0 mol% TiO2-coated samples.
electrolyte in Argon-filled glove box. Charge–discharge experi- Careful observation indicates that TiO2 particles attach non-
ments were performed galvanostatically with current density of uniformly to the particle surface. In addition, it is worthy to be
18 mA/g and 90 mA/g between 2.0 and 4.8 V on battery testers noted that TiO2 particles agglomerate to be free from parent
(Land CT2001A). The cells were cycled at 25 °C and 55 °C to material when the coating amount is over 3.0 mol%. This indi-
compare cyclic performances of the uncoated and coated samples. cates that a new phase emerges when with over 3.0 mol% of TiO2
coating, which is responsible for the diffraction peak at around
3. Results and discussion 25.5°.
3.1. Structural characterization of TiO2-coated 3.3. Differential scanning calorimetry (DSC)
Li[Li0.2Mn0.54Ni0.13Co0.13]O2
The thermal stability of various TiO2-coated Li[Li0.2Mn0.54-
The left panel of Fig. 1 describes the XRD patterns and Miller Ni0.13Co0.13]O2 electrodes was measured and the result is
indexes of various TiO2-coated Li[Li0.2Mn0.54Ni0.13Co0.13]O2 shown in Fig. 3, without removing the electrolyte. As can be
1796 J.M. Zheng et al. / Solid State Ionics 179 (2008) 1794–1799
Fig. 2. SEM images of the uncoated and TiO2-coated Li[Li0.2Mn0.54Ni0.13Co0.13]O2 samples. (a) Uncoated, (b) 1.0 mol%, (c) 3.0 mol%, (d) 5.0 mol%, (e) 7.0 mol%,
(f) 9.0 mol%.
observed, the onset and peak temperatures of thermal de- istics of the Li-ion cells is not linearly dependant, the present
composition are increased gradually by TiO2 coating. The results suggest that the thermal stability of the material at
exothermic peak temperatures of the 1.0, 3.0, 5.0 mol% TiO2- charged state is definitely improved when the material is coated
coated samples are 2.7, 7.6, and 13.3 °C, respectively, higher with TiO2. The mechanism for the improved thermal stability
than that of the uncoated sample. In addition, the heat amount is could be explained as follows. In comparison with the uncoated
also reduced considerably by TiO2 coating, viz., decreasing material, as it was discussed in Ref. [17], the TiO2 coating layer
greatly from 925 J/g for uncoated electrode to 775 J/g for plays an important role in separating the active material and the
1.0 mol%, 444 J/g for 3.0 mol% and further to 375 J/g for electrolyte and stabilizing the surface lattice oxygen. The TiO2-
5.0 mol% TiO2-coated electrodes. Although a quantitative re- coated delithiated material releases less active oxygen species
lationship between the DSC data and actual safety character- during heating process. The less the active oxygen species
J.M. Zheng et al. / Solid State Ionics 179 (2008) 1794–1799 1797
Fig. 5. Initial charge/discharge capacity and irreversible capacity loss of the
Fig. 3. DSC profiles of Li[Li0.2Mn0.54Ni0.13Co0.13]O2 cathodes at charged state
uncoated and TiO2-coated Li[Li0.2Mn0.54Ni0.13Co0.13]O2 cathodes.
to 4.8 V.
compared to the uncoated material. This is probably due to a
releases, the less the reaction occurs between the active oxygen suppression of the reaction between the cathode surface and the
species and the electrolyte remained on the material surface. electrolyte and an optimization of the solid-electrolyte interface
As a consequence, the TiO2-coated delithiated material shows (SEI) layer, which would facilitate the Li+ ion transportation
smaller thermal effect than the uncoated material. during discharge process and subsequent cycling.
The data show that the uncoated sample delivers a charge ca-
3.4. Electrochemical performance of TiO2-coated pacity of 324.5 mAh/g, the ICL of the first cycle is 75.5 mAh/g.
Li[Li0.2Mn0.54Ni0.13Co0.13]O2 By coating, the ICLs decrease with the increase of coating
amount, as shown in Fig. 5. As it was reported [12,19,20], the
Figs. 4 and 5 show the voltage/capacity curves and the initial decreases of the ICLs of the coated samples here are believed
irreversible capacity loss (ICL) in the 1st cycle of the materials due to the suppression of the side reaction between electrolyte
with different amount of surface coating. It is shown that the and the oxidative cathodes by coating layer.
first charge profiles are all accompanied with irreversible The typical cyclic performances of various TiO2-coated Li
voltage plateaux for oxidation beyond the formal oxidation [Li0.2Mn0.54Ni0.13Co0.13]O2 samples cycled at 25 °C and
states of Mn4+, Ni4+, and Co4+. The voltage plateaux have been elevated temperature (55 °C) are shown in Fig. 6(a) and (b).
assigned to an irreversible loss of oxygen from the lattice based Due to the presence of inactive TiO2 with poor electronic
on in situ X-ray diffraction studies [18]. The oxygen loss results conductivity on the electrodes, the initial discharge capacities of
in the activation of the oxidation states of Mn correspondingly the coated materials decrease with the increase of coating
at the end of first discharge. amount, no matter at 25 °C or under 55 °C. From Fig. 6(a), it is
It is noted that the 1.0 mol% and 3.0 mol% TiO2-coated found that the uncoated sample delivers 128 mAh/g after
samples deliver significantly reduced initial charge capacity 90 cycles with only 63% of capacity retention. The capacity
whereas they exhibit almost the same discharge capacity as retention of the samples coated with 1.0, 3.0, 5.0, 9.0 mol%
TiO2 are 79%, 87%, 71%, and 69%, respectively. For the
samples cycled at 55 °C, observation shows that the cyclic
performance curve for 1.0 mol% TiO2-coated sample almost
overlaps with that for uncoated sample, indicating that only
1.0 mol% TiO2 is insufficient to improve the cyclic performance
at elevated temperature. Comparison demonstrates that proper
amount of TiO2 coating clearly improve the cycling stability of
Li[Li0.2Mn0.54Ni0.13Co0.13]O2 samples, but excess amount (5.0
and 9.0 mol%) of TiO2 brings negative effect of a partial in-
sulation of the cathode particles, lowering the capacity utili-
zation and cyclic stability.
Here, attention should be paid to the charge/discharge per-
formance of the materials between 25 °C and 55 °C. It is visible
that elevating temperature up to 55 °C increases the discharge
capacity significantly. For uncoated sample, the discharge ca-
pacity in the 1st cycle increases from 203.6 mAh/g to
Fig. 4. First charge/discharge profiles of the uncoated and TiO2-coated 251.7 mAh/g, which is comparable to that of the sample cycled
Li[Li0.2Mn0.54Ni0.13Co0.13]O2 cathodes. at 18 mA/g over the same potential range. It suggests that
1798 J.M. Zheng et al. / Solid State Ionics 179 (2008) 1794–1799
the conductivity of Li[Li0.2Mn0.54Ni0.13Co0.13]O2 cathode is
relatively low, elevating temperature could enhance the
conductivity of the material and, therefore, increase the dis-
charge capacity. However, cyclic performances of the uncoated
sample under the two temperatures degrade at almost the same
rate. By coating, in particular, 3.0 mol% of TiO2, not only the
discharge capacity is increased by some 40 mAh/g, but also the
cyclic performance is improved greatly. This manifests that it's
possible to use this compound under elevated temperature
(55 °C) at medium C-rate to gain high capacity, meanwhile, and
still need to overcome the drawback of low conductivity of the
material in the further research.
3.5. XPS spectra of the TiO2-coated Li[Li0.2Mn0.54Ni0.13Co0.13]O2
cathodes Fig. 7. XPS spectra of the uncoated and 3.0 mol% TiO 2 -coated
Li[Li 0.2 Mn 0.54 Ni 0.13 Co 0.13 ]O 2 cathodes.
To understand the effects of TiO2 coating on the electro-
chemical performance of the Li[Li0.2 Mn0.54Ni0.13Co0.13]O2
cathode materials, XPS spectra of C 1s and F 1s of the uncoated Analysis of the F 1s chemical state reveals the presence of
and 3.0 mol% TiO2-coated samples are compared in Fig. 7(a) two dominant binding energies (BEs), one at around 685 eV and
and (b), respectively. Since neither the binder of PVDF nor the other at about 687.5 eV. The former peak can be assigned to
conductive acetylene black is contained in the electrode, the LiF, while the latter one can be assigned to Li xPO yF z [23]. It is
peak around 285 eV results from the peaks of the carbon of obvious that the contents of LiF and Li xPO yF z on the surface of
polyethers, alcoholic and methylene groups [21]. The shoulder coated sample are both much lower than that on the surface of
peak at about 286.5 eV and the peak centered at around the uncoated sample. LiPF6 is known to be chemically and
288.5 eV are assigned to carbon atom surrounded by one oxy- thermodynamically unstable and its reaction product HF could
gen atom and the one in a three-oxygen environment, respec- react with Li2CO3 to form LiF. Actually, LiPF6-based elec-
tively. These originate from carbonate groups involved in trolyte always contains a small amount of water. LiPF6 decom-
polymerized carbonates formed by direct polymerization of EC poses thermodynamically during cycling and reacts with water
and initiated by EC oxidation. The direct polymerization of EC to form HF acid. The HF can react with carbonates like Li2CO3
by strong Lewis acid, e.g., PF5, into a poly(oxyethylene-alt- to form LiF as suggested by Aurbach et al. [24] and EdstrÖm
ethylene carbonate) is possible [22]. As observed, the peaks et al. [25].
around 286.5 eV and 288.5 eV are both weaker for 3.0 mol% Based on this, it is convinced that the coating layer sup-
TiO2-coated sample than that for the uncoated sample. It sug- presses the oxidation of the solvent and mitigates the formation
gests that the polymerization and oxidation of EC and/or DMC of LiF on the surface. Because LiF film in the cathode surfaces
are suppressed by TiO2 coating. This is one reason why the is insulating for lithium-ion transport [22], the 3.0 mol% TiO2-
3.0 mol% TiO2-coated sample renders better electrochemical coated material with lower content of LiF in their surfaces
performance. displays much lower electrode polarization after a few cycles
and exhibits much better electrochemical performance of cyc-
ling stability than the uncoated material.
Fig. 6. Cycle-life performances of the uncoated and TiO2-coated Li[Li0.2Mn0.54
Ni0.13Co0.13]O2 cathodes at 90 mA/g, over the potential range 2.0~4.8 V. (a) Fig. 8. Voltage/capacity curves of the uncoated and 3.0 mol% TiO2-coated
25 °C; (b) 55 °C. Li[Li0.2Mn0.54Ni0.13Co0.13]O2 cathodes.
J.M. Zheng et al. / Solid State Ionics 179 (2008) 1794–1799 1799
Fig. 8 compares the Voltage/capacity curves of the uncoated 2007CB209702), National Natural Science Foundation of
and 3.0 mol% TiO2-coated Li[Li0.2Mn0.54Ni0.13Co0.13]O2 in 1st China (NNSFC, Grant No. 20433060, 29925310).
and 60th cycles cycled at 90 mA/g and at 25 °C. It is seen that
the polarization of sample with 3.0 mol% TiO2 in the 1st dis- References
charge process is larger than that of uncoated sample, exhibit-
ing a larger difference between charge and discharge plateaux. [1] Z. Lu, D.D. MacNeil, J.R. Dahn, Electrochem. Solid-State Lett. 4 (2001)
A191.
However, after 60 cycles, the polarization of uncoated sample
[2] Z. Lu, J.R. Dahn, et al., J. Electrochem. Soc. 149 (2002) A778.
becomes larger than that of 3.0 mol% TiO2-coated sample. [3] D.D. MacNeil, Z. Lu, J.R. Dahn, J. Electrochem. Soc. 149 (2002) A1332.
Among the many factors, resistance increase of the SEI layer [4] Z. Lu, J.R. Dahn, J. Electrochem. Soc. 149 (2002) A1454.
may play an important role in the changes of polarization of [5] Y. Wu, et al., Electrochem. Solid-State Lett. 9 (5) (2006) A221.
charge/discharge plateaux for these materials. [6] J. Cho, Y.J. Kim, B. Park, Chem. Mater. 12 (2003) 3788.
[7] J. Cho, Y.J. Kim, et al., Angew. Chem., Int. Ed. Engl. 40 (2001) 3367.
[8] Z.H. Chen, J.R. Dahn, Electrochem. Solid-State Lett. 5 (2002) A213.
4. Conclusion [9] J. Cho, C.-S. Kim, S.-I. Yoo, Electrochem. Solid-State Lett. 3 (2000) 362.
[10] M. Mladenov, R. Stoyanova, E. Zhecheva, S. Vassilev, Electrochem.
The Li[Li0.2Mn0.54Ni0.13Co0.13]O2 samples coated with TiO2 Commun. 3 (2001) 410.
have been prepared and investigated. Due to TiO2 coating, the [11] S.M. Lee, S.H. Oh, J.P. Ahn, W. Il Cho, H. Jang, J. Power Sources 159 (2)
onset temperature of thermal decomposition was increased (2006) 1334.
[12] H.S. Liu, Z.R. Zhang, Z.L. Gong, Y. Yang, Solid State Ion. 166 (2004) 321.
and the amount of heat for the exothermic peak was reduced [13] B.V.R. Chowdari, G.V.S. Rao, S.Y. Chow, J. Solid State Electrochem. 6
significantly. The ICLs in the 1st cycle decreased with the (2002) 565.
increase of the coating amount of TiO2. The significantly [14] G.T.K. Fey, P. Muralidharan, C.Z. Lu, Y.D. Cho, Solid State Ion. 176 (2005)
improved cycle-life performance was also observed in the 2759.
3.0 mol% TiO2-coated Li[Li0.2Mn0.54Ni0.13Co0.13]O2 samples. [15] Y.-J. Kang, J.-H. Kim, S.-W. Lee, Y.-K. Sun, Electrochim. Acta 55 (2005)
4784.
The 3.0 mol% TiO2-coated Li[Li0.2Mn0.54Ni0.13Co0.13]O2 [16] J. Li, J.M. Zheng, Y. Yang, J. Electrochem. Soc. 154 (5) (2007) A427.
sample displayed pronouncedly an improved capacity retention [17] Z.R. Zhang, J. Li, Y. Yang, Electrochim. Acta 52 (2006) 1442.
at elevated temperature (55 °C). It is proposed that the coating [18] Z.H. Lu, J.R. Dahn, J. Electrochem. Soc. 149 (7) (2002) A815.
layer avoided the direct contact between the active material and [19] Z.R. Zhang, H.S. Liu, Z.L. Gong, Y. Yang, J. Power Sources 129 (2004)
the electrolyte and thus mitigated a side reaction between them. 101.
[20] G.T.K. Fey, C.Z. Lu, T.P. Kumar, Y.C. Chang, Surf. Coat. Technol. 199
This decreased the amount of the product from the polymeriza- (2005) 22.
tion and oxidation of EC and reduced the formation LiF. These [21] Y.K. Fan, J.M. Wang, Z. Tang, W.C. He, J.Q. Zhang, Electrochim. Acta 52
were responsible for better electrochemical performance of cycl- (2007) 3870.
ing stability of the TiO2-coated Li[Li0.2Mn0.54Ni0.13Co0.13]O2 [22] K. Edstrom, T. Gusafuuon, J.O. Thomas, Electrochim. Acta 55 (2004) 397.
[23] W.T. Li, B.L. Lucht, J. Electrochem. Soc. 153 (8) (2006) A1617.
cathode materials.
[24] B. Markovsky, A. Rodkin, et al., J. Electrochem. Soc. 151 (2004) A1068.
[25] K. EdstrÖm, T. Gustafsson, J.O. Thomas, Electrochim. Acta 55 (2004)
Acknowledgement 379.
This work was financially supported by the National Basic
Research Program of China (973 Program) (Grant No.