Materials Chemistry and Physics 117 (2009) 41–45

Contents lists available at ScienceDirect

Materials Chemistry and Physics

journal homepage: www.elsevier.com/locate/matchemphys

Effects of different carbonate precipitators on LiNi1/3 Co1/3 Mn1/3 O2 morphology

and electrochemical performance
Haibo Ren a , Yanghui Huang a,b , Yunhong Wang a , Zhongjing Li a , Ping Cai a ,
Zhenghe Peng a,∗ , Yunhong Zhou a
a
Department of Chemistry, Wuhan University, Wuhan 430072, PR China
b
Department of Applied Chemistry, East China Institute of Technology, Fuzhou 344000, PR China

a r t i c l e i n f o a b s t r a c t

Article history: Compounds LiNi1/3 Co1/3 Mn1/3 O2 are prepared by adopting different carbonate, i.e., NaHCO3 , NH4 HCO3
Received 3 September 2008 and Na2 CO3 under the same conditions, which results in the impressive differences in the morphology
Received in revised form 18 December 2008 properties and electrochemical behaviors of LiNi1/3 Co1/3 Mn1/3 O2 . The microscopic structural features of
Accepted 2 January 2009
LiNi1/3 Co1/3 Mn1/3 O2 are investigated using X-ray diffraction (XRD) and Scanning Electron Microscopy
(SEM). The electrochemical properties of LiNi1/3 Co1/3 Mn1/3 O2 are carried out by charge–discharge cycling
Keywords:
tests and the electrochemical reactions during charge–discharge process are also investigated by the cyclic
Composite materials
voltammetry (CV). Experimental results show that the microscopic structural features, the morphology
Sintering
Crystal structure
properties and electrochemical performances are deeply related to the precipitators. In addition, the cyclic
Electrochemical properties voltammetry indicates that the oxidation of nickel occurs by two steps Ni2+ → Ni3+ → Ni4+ .
© 2009 Published by Elsevier B.V.

1. Introduction particle morphology. As a result, it is difficult to prepare this com-

plicated material, and the rate capability of this material depends
Lithium-ion batteries have presently become attractive for on the synthetic route which has greater influence on the particle
portable electronic devices because of their higher output volt- morphology. Therefore, it is important for LiNi1/3 Co1/3 Mn1/3 O2 to
age and energy density than other rechargeable systems. Originally, select suitable preparation method. In order to synthesize uniform
commercial lithium-ion cells employ the layered LiCoO2 cathode, and homogeneous LiNi1/3 Co1/3 Mn1/3 O2 , some research groups have
which has a theoretical capacity of 275 mAh g−1 . The practically reported hydroxide coprecipitation method [10,15–17]. In fact,
attainable capacity, however, is found to be only 120–130 mAh g−1 in hydroxide coprecipitation method, Mn(OH)2 (Mn2+ ) oxidizes
in the voltage range 2.7–4.2 V [1,2]. To increase the capacity, several gradually to MnOOH (Mn3+ ) or MnO2 (Mn4+ ) upon precipitation
cationic substitutions such as LiNi0.8 Co0.2 O2 [3,4], LiNi0.5 Mn0.5 O2 conditions which can decrease the homogeneity of the final prod-
[5,6] and LiNi1/3 Co1/3 Mn1/3 O2 [7–12] are being intensively pur- uct. Therefore, it is critical to control the valence state of Mn stable in
sued. Especially, computation [13] and XPS [14] studies on aqueous solution. The carbonate coprecipitation method adopted to
LiNi1/3 Co1/3 Mn1/3 O2 show that the predominant oxidation states prepare the composite precursor could keep valence state of Mn2+
of Ni, Co, Mn ions are +2, +3, +4, respectively. In such a system, stable in aqueous solution during precipitation process so that the
the electrochemical capacity is derived predominantly from redox homogeneous solid solution could be obtained. In our work, the
reactions on the nickel and cobalt ions. The relatively high con- NaHCO3 , NH4 HCO3 and Na2 CO3 are all employed as precipitator to
centration of inactive Mn4+ ions that remains constant throughout yield LiNi1/3 Co1/3 Mn1/3 O2 , of which differences in structure, mor-
charge and discharge is believed to contribute to the enhanced sta- phology and electrochemical properties are also reported.
bility and higher reversible capacity of these types of electrodes
compared to those containing only cobalt and nickel. Thus, this solid 2. Experimental
solution could be one of the promising cathode materials. Unfortu-
nately, due to the complicated composition, the cathode properties LiNi1/3 Co1/3 Mn1/3 O2 was synthesized by carbonate coprecipitation method.
Three of aqueous solutions (50 ml) containing NiSO4 ·6H2 O, CoSO4 ·7H2 O and
of LiNi1/3 Co1/3 Mn1/3 O2 should be sensitively dependent on micro-
MnSO4 ·H2 O (cationic ratio of Ni:Co:Mn = 1:1:1) with a total metal ion concentration
scopic features like the distribution of transition metal and the of 0.5 M were dropped into the strongly stirred 50 ml of 1 M NaHCO3 , NH4 HCO3
and Na2 CO3 solutions at a speed of one drop per second at 80 ◦ C, respectively. After
reaction for 10 h, the sediment was filtered with distilled water for several times
before being dried at 120 ◦ C for 5 h. The obtained precursor powder was mixed with
∗ Corresponding author. Tel.: +86 27 87218254; fax: +86 27 68754067. 6% excess Li2 CO3 (excess amount of Li salt was used to compensate for possible Li
E-mail address: [email protected] (Z. Peng). loss during the calcination) by a mortar and the powder was pressed into pellets.

0254-0584/$ – see front matter © 2009 Published by Elsevier B.V.

doi:10.1016/j.matchemphys.2009.01.002
42 H. Ren et al. / Materials Chemistry and Physics 117 (2009) 41–45

be used to identify the cation mixing degree. Generally, when the

value of intensity ratio is more than 1.2, the cation mixing is small
with good layered structure and the higher value indicates lower
cation mixing. According to Table 1, the intensity ratios of the sam-
ple A, B and C are 1.74, 1.62 and 1.60, respectively. The sample A
prepared by NaHCO3 shows the highest structural integrity, i.e.,
lower cation mixing, whereas the sample C by Na2 CO3 shows the
lowest one. After making a comparison of the values of c/a and
those of I(0 0 3) /I(1 0 4) , it could be guessed that the sample A should
have the best crystal structure and will show better electrochemical
properties.
Normally, the crystal structure and morphology are two main
factors which affect the electrochemical performance of the mate-
rials. Fig. 2 shows the SEM micrographs of the three samples with
different magnifications. The powder of the sample A and B is
quasi-spherical secondary particle formed by the primary particle
(<1 ␮m) aggregating each other. This kind of morphology could be
thought to enhance high rate capability with respect to high sur-
face area without decreasing of tapping density [22]. From Fig. 2a
Fig. 1. XRD patterns of LiNi1/3 Co1/3 Mn1/3 O2 prepared from different carbonate.
and c, we can see that the particle size of the sample A has a bet-
ter consistency than that of the sample B. The uniform distribution
The pellets were initially heated to 500 ◦ C for 6 h in air and reground later. The leads to the uniform depth of charge (DOC) of each particle, which
pellets were remade and subsequently calcined at 850 ◦ C for 20 h in air to obtain increases the utilization of the material to enhance the overall bat-
LiNi1/3 Co1/3 Mn1/3 O2 . The LiNi1/3 Co1/3 Mn1/3 O2 samples from NaHCO3 , NH4 HCO3 and tery performances. The powder of the sample C is formed by the
Na2 CO3 were denoted as sample A, sample B and sample C, respectively.
primary particle, but these small particles are aggregated into irreg-
X-ray diffraction measurements of the LiNi1/3 Co1/3 Mn1/3 O2 materials were
carried out using X-ray diffraction (Shimadzu XRD-6000) with Cu K␣ radiation ular shapes. We think that this kind of morphology does not benefit
( = 1.54056 Å). Lattice parameters and unit-cell volumes were calculated by a least its electrochemical performance.
squares method with FullProf Suite program. Particle morphology of the powders Testing cells are firstly operated at a current density of 50 mA g−1
after calcination was observed using a scanning electron microscope (SEM, QUANTA- in the voltage range of 3–4.3 V. The initial charge–discharge curves
200). Electrochemical charge–discharge experiments were performed using the
of the three samples are presented in Fig. 3, and their cycling per-
CR2016 coin-type cell. Test cathode electrodes were prepared by mix 80:15:5 (mass
ratio) of active material, acetylene black and PTFE binder, respectively, in isopropyl formances under this condition are offered in Fig. 4. As shown in
alcohol. The test cells were assembled with the electrode prepared above as cath- Fig. 3, the first charge/discharge specific capacities are 187.6/166.6,
ode, lithium metal as anode, and Celgard 2300 film as separator in an argon-filled 189.8/164.3 and 188.8/162.9 mAh g−1 with a coulombic efficiency
glove box. The electrolyte was 1 M LiPF6 dissolved in EC + DMC (1:1 volume ratio).
about 88.8%, 86.6% and 86.3% for samples A, B and C, respec-
Cyclic voltammetry (CV) experiments were carried out at a scanning rate 0.1 mV s−1
between 2.8 and 4.8 V versus Li/Li+ . All tests were performed at room temperature.
tively. All cells show coulombic inefficiencies; i.e., more capacity
is observed upon charge than on discharge. The irreversible capac-
ity loss might be attributed to the formation of a solid electrolyte
3. Results and discussion
interface (SEI) on the surface of the electrode and insufficient
soaking of the electrode material during the first cycle [23]. More-
Fig. 1 illustrates the powder X-ray diffraction patterns of the
over, we can also find from Fig. 3 that the polarization degree of
samples A, B and C. The XRD patterns of all samples could be
the electrode composed of sample A, B and C exists difference in
indexed based on the ␣-NaFeO2 structure (space group: R3̄m) with-
charge/discharge process. The electrode composed of the sample A
out any impurity peaks. The lattice parameters of the samples were
shows a lower polarization and the electrode composed of the sam-
calculated by least square method using 10-diffraction lines. The
ple C has a higher polarization, which may be one of the reasons
determined parameters are summarized in Table 1. The lattice
leading to the different charge/discharge capacities. According to
parameters are somewhat different for the samples. The sample
Fig. 4, at the end of the 30th cycle, the retained discharge capaci-
A shows the smallest lattice parameter a and the biggest lattice
ties for the sample A, B and C are 158.8, 153.6 and 151.6 mAh g−1 ,
parameter c, thus it holds a larger triangle distortion, c/a. In addi-
respectively, which are 95.3%, 93.5% and 93.1% of initial discharge
tion, the three samples have smaller lattice parameters, a and c
capacity.
compared to those reported [14,18] while a larger triangle distor-
Obviously, the sample A synthesized from NaHCO3 reveals
tion, c/a, but it is similar to some results of Reddy et al. [19,20].
excellent electrochemical performance. The sample C, However,
Reddy et al. [19] considered these smaller lattice parameters might
synthesized from Na2 CO3 shows poor electrochemical properties.
be due to smaller cation mixing and better ordering of the transi-
These results are consistent with the analysis of XRD and SEM.
tion metal ions in the metal-layer. Hwang et al. [21] believed that
LiNi1/3 Co1/3 Mn1/3 O2 is only partially de-intercalated and inter-
the intensity ratio of the (0 0 3) and (1 0 4) peaks (I(0 0 3) /I(1 0 4) ) could
calated when cells are charged and discharged in the voltage range
of 3–4.3 V. Higher discharge capacities can be obtained when the
Table 1 voltage range is broadened, but the increased coulombic inefficien-
Lattice parameters of LiNi1/3 Co1/3 Mn1/3 O2 prepared by different carbonate. cies and more rapid capacity fading will be brought. To illustrate
these phenomena and show the differences of the three samples
Samples a (Å) c (Å) c/a I(0 0 3) /I(1 0 4)
operated in broader voltage, the voltage range 2.5–4.6 V (50 mA g−1 )
A 2.8478 14.1794 4.98 1.74 is adopted. Here, 4.6 V is chosen as the upper limit because the criti-
B 2.8485 14.1552 4.97 1.62
C 2.8542 14.1759 4.97 1.60
cal upper limit of LiNiO2 and LiCoO2 is 4.6 V [18,24]. Fig. 5 shows the
Literature [10] 2.864 14.233 4.969 0.8 cycling performance of the sample A, B and C under given condi-
Literature [13] 2.867 14.246 4. 969 – tions and Table 2 also gives some data about the charge–discharge
Literature [14] 2.8467 14.1967 4.99 1.6 capacities, coulombic efficiency and capacity retention rations of
Literature [15] 2.8496 14.2045 4.98 1.87
the three samples under the same conditions. For comparison pur-
 H. Ren et al. / Materials Chemistry and Physics 117 (2009) 41–45 43

Fig. 2. SEM micrographs of LiNi1/3 Co1/3 Mn1/3 O2 prepared from different carbonate. (a and b) The sample A; (c and d) the sample B; (e and f) the sample C.

Table 2
The charge–discharge data for the samples A, B and C in the voltage range of 2.5–4.6 V.

Samples 1st cycle Qcharge 1st cycle Qdischarge 1st cycle coulombic efficiency (%) 30th cycle Qdischarge 30th cycle capacity retention (%)
(mAh g−1 ) (mAh g−1 ) (mAh g−1 )
2.5–4.6 V 3–4.3 V 2.5–4.6 V 3–4.3 V

A 228.4 200.5 87.8 88.8 177.0 88.3 95.3

B 228.9 196.2 85.7 86.6 175.2 89.3 93.5
C 230.1 193.8 84.2 86.3 158.1 81.6 93.1
44 H. Ren et al. / Materials Chemistry and Physics 117 (2009) 41–45

Fig. 3. Initial charge–discharge curves of the three samples operated in the voltages
range of 3–4.3 V under a current density 50 mA g−1 .

Fig. 4. Specific discharge capacity as a function of cycle number for the samples
operated in the voltage range of 3–4.3 V under a current density 50 mA g−1 .

Fig. 6. CV curves of the three samples. (a) The sample A; (b) the sample B; (c) the
sample C.

poses, some data from the narrower voltage range are also listed in
Table 2. From the table, we can find that the initial discharge capaci-
ties for the three samples are remarkably increased. However, their
coulombic efficiencies decrease and the speed of capacity fading
increases. In addition, from Figs. 4 and 5, it is noticed that the sample
C has different cycling characters under the different voltage range.
The discharge capacity experiences three stages, which are decline,
stability and decline again under the narrower voltage range. How-
ever, the discharge capacity descends linearly without a stable stage
under the broader voltage range. These phenomena indicate that
the different charge–discharge voltage range may have different
influence on the electrochemical performances of electrode mate-
Fig. 5. Specific discharge capacity as a function of cycle number for the samples rial. Essentially, it should be attributed to the different changes of
operated in the voltage range of 2.5–4.6 V under a current density 50 mA g−1 . microstructure under the different charge–discharge voltage range.
 H. Ren et al. / Materials Chemistry and Physics 117 (2009) 41–45 45

We carry out cyclic voltammetry to monitor the electrochemi- References

cal reactions during charge–discharge process as shown in Fig. 6.
Fig. 6a–c shows only the first cycle curves of the sample A, B and [1] A. Manthiram, J. Kim, Chem. Mater. 10 (1998) 2895–2909.
[2] J. Fan, P.S. Fedkiw, J. Power Sources 72 (1998) 165–173.
C, respectively. All the three samples show the two anodic peaks at [3] G.T.K. Fey, R.F. Shiu, V. Subramanian, J.G. Chen, C.L. Chen, J. Power Sources 103
3.6–3.9 and 4.6–4.8 V whereas the corresponding cathodic peak is at (2002) 265–272.
3.5–3.8 and 4.5–4.7 V. They are ascribed to Ni2+ /Ni4+ and Co3+ /Co4+ [4] S.B. Yi, H.T. Chung, H.G. Kim, Electrochem. Commun. 9 (2007) 591–595.
[5] D.H. Park, S.T. Lim, S.J. Hwang, C.S. Yoon, Y.K. Sun, J.H. Choy, Adv. Mater. 17 (2005)
redox couples [14], respectively. However, an anodic peak, around 2834–2837.
3.9–4.4 V, can be obviously observed in Fig. 6a and c. It should be [6] Y. Hinuma, Y.S. Meng, K. Kang, G. Ceder, Chem. Mater. 19 (2007) 1790–1800.
assigned as the oxidation of Ni3+ → Ni4+ [25,26]. So we think the [7] H. Ren, Y. Wang, D. Li, L. Ren, Z. Peng, Y. Zhou, J. Power Sources 178 (2008)
439–444.
oxidation of nickel occurs by two steps Ni2+ → Ni3+ → Ni4+ rather [8] S. Venkatraman, J. Choi, A. Manthiram, Electrochem. Commun. 6 (2004)
than in a single step Ni2+ → Ni4+ . Each oxidation process in charge 832–837.
process could be assigned as Ni2+/3+ (3.6–3.9 V) [20,21], Ni3+/4+ [9] J. Li, J.M. Zheng, Y. Yang, J. Electrochem. Soc. 154 (2007) A427–A432.
[10] C. Deng, L. Liu, W. Zhou, K. Sun, D. Sun, Electrochim. Acta 53 (2008)
(3.9–4.4 V) and Co3+/4+ (4.6–4.8 V), respectively.
2441–2447.
[11] P. Reale, D. Privitera, S. Panero, B. Scrosati, Solid State Ion. 178 (2007)
4. Conclusions 1390–1397.
[12] H.Y. Chang, C.L. Sheu, S.Y. Cheng, J. Power Sources 174 (2007) 985–989.
[13] B.J. Hwang, Y.W. Tsai, D. Carlier, G. Ceder, Chem. Mater. 15 (2003)
Layered compounds LiNi1/3 Co1/3 Mn1/3 O2 are prepared by dif- 3676–3682.
ferent carbonate coprecipitation methods, which results in the [14] K.M. Shaju, G.V. Subba Rao, B.V.R. Chowdari, Electrochim. Acta 48 (2002)
differences in morphology and microstructure by the SEM observa- 145–151.
[15] B. Lin, Z. Wen, Z. Gua, S. Huang, J. Power Sources 175 (2008) 564–569.
tion and the XRD experiment. The electrochemical tests show that [16] X. Luo, X. Wang, L. Liao, S. Gamboa, P.J. Sebastian, J. Power Sources 158 (2006)
compound LiNi1/3 Co1/3 Mn1/3 O2 synthesized from NaHCO3 holds 654–658.
the higher initial discharge capacities of 166.6 mAh g−1 (3–4.3 V, [17] S. Zhang, Electrochim. Acta 52 (2007) 7337–7342.
[18] T. Ohzuku, Y. Makimura, Chem. Lett. 7 (2001) 642–643.
50 mA g−1 ) and 200.5 mAh g−1 (2.5–4.6 V, 50 mA g−1 ), as well as [19] M.V. Reddy, G.V. Subba Rao, B.V.R. Chowdari, J. Power Sources 159 (2006)
good cycling performance, which are superior to those of the mate- 263–267.
rials from NH4 HCO3 and Na2 CO3 . Cyclic voltammetry results show [20] J. Guo, L.F. Jiao, H.T. Yuan, H.X. Li, M. Zhang, Y.M. Wang, Electrochim. Acta 51
(2006) 3731–3735.
the oxidation of nickel occurs by two steps Ni2+ → Ni3+ → Ni4+ [21] B.J. Hwang, R. Santhanam, C.H. Chen, J. Power Sources 114 (2003) 244–252.
rather than in a single step Ni2+ → Ni4+ . [22] T.H. Cho, S.M. Park, M. Yoshio, T. Hiraib, Y. Hideshima, J. Power Sources 142
(2005) 306–312.
[23] Z. Wang, Y. Sun, L. Chen, X. Huang, J. Electrochem. Soc. 151 (2004)
Acknowledgement A914–A921.
[24] J.M. Paulsen, C.L. Thomas, J.R. Dahn, J. Electrochem. Soc. 147 (2000) 861–868.
We gratefully acknowledge the financial support from the [25] J. Reed, G. Ceder, Electrochem. Solid State Lett. 5 (2002) A145–A148.
[26] Y.S. Hong, Y.J. Park, K.S. Ryu, S.H. Chang, Y.J. Shin, J. Power Sources 147 (2005)
National Nature Science Foundation of China (No. 29833090, No. 214–219.
29771025 and No. 20573078).