www.elsevier.com/locate/elecom

Electrochemistry Communications 9 (2007) 965–970

Significant improved electrochemical performance ofLi(Ni1/3Co1/3Mn1/3)O2 cathode on volumetric energy density

and cycling stability at high rate

Yongguang Liang a,*, Xiaoyan Han a, Xinwen Zhou b, Jutang Sun a,*, Yunhong Zhou a

a Department of Chemistry, Wuhan University, Wuhan 430072, PR Chinab State Key Laboratory of Physical Chemistry of Solid Surfaces, Xiamen University, Xiamen 361005, PR China

Received 19 October 2006; received in revised form 24 November 2006; accepted 30 November 2006Available online 9 January 2007

Abstract

Li(Ni1/3Co1/3Mn1/3)O2 microspheres with a tap density of 2.41 g cm�3 have been synthesized for applications in high power and highenergy systems, using a simple rheological phase reaction route. Cyclic voltammograms (CV) showed no shift of anodic and cathodicpeaks centred at 3.81, 3.69 V for the Ni2+/Ni4+ couple after first cycle. The results of power pulse area specific impedance (ASI) anddifferential scanning calorimetry (DSC) tests showed lower power impedance and increased thermal stability of the electrode at high rate.These merits mentioned above provided significant improved capacity and rate performance for Li(Ni1/3Co1/3Mn1/3)O2 microspheres,which 159, 147 mAh g�1 discharge capacity was delivered after 100 cycles between 2.5–4.6 V vs. Li at a different discharge rate of2.5 C (500 mA g�1), 5 C and a constant 0.5 C charge rate, respectively.� 2006 Elsevier B.V. All rights reserved.

Keywords: Li(Ni1/3Co1/3Mn1/3)O2; Electrochemical performance; High rate; Volumetric energy density; Cycling stability; Lithium-ion batteries

1. Introduction

LiCoO2 has been used as a major cathode material forlithium-ion batteries since it was firstly introduced by SonyCorp. [1]. Obviously, it is an excellent cathode materialwith low irreversible capacity loss and good cycling perfor-mance. However, the newly developed electric vehicles(EV) and hybrid electric vehicles (HEV) need high volu-metric energy density over LiCoO2 and cobalt price becamemore expensive nowadays. These drawbacks of LiCoO2

accelerate intensive study to find an alternative cathodematerial for high energy applications and the practicalcapacity of lithium-ion battery has been increased gradu-ally for cathode material to meet the requirements ofhigh-power and high-energy applications [2–4].

1388-2481/$ - see front matter � 2006 Elsevier B.V. All rights reserved.

doi:10.1016/j.elecom.2006.11.039

* Corresponding authors. Tel.: +86 27 87218494; fax: +86 27 68754067.E-mail address: ygliang@mail.whu.edu.cn (Y. Liang).

Recently, a layered transition metal oxide with hexago-nal structure, Li(Ni1/3Co1/3Mn1/3)O2, was introduced byOhzuku and Makimura as a candidate cathode materialto replace LiCoO2 [5]. This material attracts significantinterest because the combination of nickel, manganeseand cobalt can provide advantages such as a reversiblecapacity of more than 200 mAh g�1 within the voltagerange of 2.5–4.6 V (theoretical value is 278 mAh g�1),milder thermal stability at charged state, lower cost andless toxicity than LiCoO2 [5–8]. Unfortunately, it is difficultto prepare this compound and this material synthesizedusing traditional solid-state or solution methods couldshow low rate capability and inferior electrochemical per-formance (lower capacity, poor cycling and rate perfor-mance) due to the resulted inhomogeneous or impurephase [5,9–12]. Although the hydroxide or carbonate co-precipitation method helps to obtain superior electrodeperformance, such an improvement is achieved at thecareful control of the experimental conditions, e.g., pH,

Fig. 1. Rietveld analysis of the powder XRD patterns for sphericalLi(Ni1/3Co1/3Mn1/3)O2 synthesized by a rheological phase reaction route,using Fullprof suite program.

966 Y. Liang et al. / Electrochemistry Communications 9 (2007) 965–970

temperature, and mixing speed in order to obtain a homo-geneous hydroxide or carbonate precipitate [14–17]. Thepreparation procedures are complex and not suitable fora large scale up [9–23]. Therefore, it is important to searchfor a simpler route to synthesize Li(Ni1/3Co1/3Mn1/3)O2

cathode material with improved rate performance and vol-umetric energy density before it can be used in practicalcells.

In this paper, we introduced a novel method to preparespherical Li(Ni1/3Co1/3Mn1/3)O2 with high volumetricenergy density by a simple rheological phase reactionroute. The electrochemical performance of Li/Li(Ni1/3-Co1/3Mn1/3)O2 cell at high rate has also been investigated.

2. Experimental

Synthesis of high-tap-density Li(Ni1/3Co1/3Mn1/3)O2

was carried out using a rheological phase reaction method.LiOH Æ H2O, amorphous MnO2, b-Ni(OH)2 and Co2O3

were employed as the starting materials in quantities corre-sponding to 0.1 mol stoichiometric Li(Ni1/3Co1/3Mn1/3)O2

with a 4 mol% excess of Li. The chemicals were fully mixedby grinding and 2 mL water was added to get a rheologicalbody. The mixture was transferred into a container and thecontainer was then sealed in a stainless autoclave at 80 �Cfor 8 h to give a grey precursor. The precursor was treatedat 930 �C in air for 12 h to yield Li(Ni1/3Co1/3Mn1/3)O2.Traditional solid state reaction method (denoted as SP)and co-precipitation method (denoted as CP) were alsoused to synthesize Li(Ni1/3Co1/3Mn1/3)O2 as a contrast torheological phase reaction method reported in this work(denoted as RP).

Identification of phases and structures were performedon a Bruker AXS D8 diffractometer with monochromaticCu Ka1 radiation (1.54056 A), using a step-scan mode witha step of 0.01� in the 2h range of 10�–120�. Chemical anal-ysis was carried out by inductively coupled plasma-atomicemission spectroscopy (ICP-AES, model IRIS, TJA). Themorphological features were observed by scanning electronmicroscope (FEI QUANTA 200). Particle size distributionwas determined by low-angle laser light scattering(LALLS) measurements (Malvern Mastersizer 2000). TheJZ-1 tap density tester was used to detect tap density dataof three samples. A CHI 760 C workstation was alsoapplied for cyclic voltammogram test using a scan rate of0.05 mV S�1. The thermal stability of Li(Ni1/3Co1/3Mn1/3)-O2 electrodes was examined by means of DSC with a Net-zsch STA 449 thermal analysis system from 100 to 350 �Cat a heating rate of 5 �C min�1.

The galvanotactic cycling was conducted using a coin-type cell (size 2016), which consisted of a working electrodeand a lithium foil counter electrode separated by a Celgard2400 microporous membrane. Positive electrode wasprepared by pasting the mixture of 80% Li(Ni1/3Co1/3-Mn1/3)O2 powders, 15% acetylene black and 5% PTFEonto a aluminum foil current collector. A 1 mol/L solutionof LiPO6 dissolved in EC/DEC (1:1, volume) was used as

the electrolyte. The cells were assembled in an Argon-filledglove box (Mikrouna Super 1220/750). Hybrid pulse powercharacterization (HPPC) tests were carried out at roomtemperature using the method described in literature [24].The cells of three samples were fabricated with 15.5 cm2

electrode area in a 32 cm2 stainless steel fixture. After tenformation cycles at a slow rate of 0.1 C (20 mA g�1), thesecells were charged to a certain state of charge (SOC) andthen disassembled in the glove box. Two symmetric cells(size 2016) were made out of each disassembled 15.5 cm2

electrode.

3. Results and discussion

Powder XRD patterns of Li(Ni1/3Co1/3Mn1/3)O2 samplesynthesized using a rheological phase reaction method arepresented in Fig. 1. Since this is a new synthesis, Rietveldrefinement was carried out to verify homogeneous Li(Ni1/3-Co1/3Mn1/3)O2 material had been prepared. This modelon the structure of LiCoO2 (a-NaFeO2 type) provided anexcellent fit to the data, as can be seen in Fig. 1, corre-sponding to a Robs = 5.33%, Rp = 7.24 and Rwp = 8.99.Refined unit-cell parameters, in a hexagonal setting, werea = 2.8636(2) and c = 14.2472(1) A. Chemical analysis con-firmed the composition Li0.99(Ni0.335Co0.333Mn0.332)O2

within experimental error. The nicely split of (006,102)and (108, 110) peak pairs in the XRD patterns reveal a lay-ered nature of the compound.

Particle sizes and morphological features were examinedby scanning electron microscopy, as shown in Fig. 2a. Wellsphere-shaped particles are distributed in a narrow diame-ter range of 3–7 lm. Fig. 2b gives the particle-size distribu-tion in terms of percentage of total particle volume vs.particle diameter of Li(Ni1/3Co1/3Mn1/3)O2 sample (RP),as estimated by LALLS measurements. The distributionpresents mono population of particles with size rangingfrom 0.49 to 11 lm. The main population (�70%) is

Fig. 2. Particle size and morphological features of Li(Ni1/3Co1/3Mn1/3)O2

powders prepared by rheological phase reaction route: (a) SEM image;(b) Particle size distribution vs. percentage of particle volume determinedby LALLS measurements.

Fig. 3. Electrochemical behaviors of Li(Ni1/3Co1/3Mn1/3)O2 microspheresobtained by rheological phase reaction method: (a) Charge/dischargecurves of Li/Li(Ni1/3Co1/3Mn1/3)O2 cell at rate of 0.5 C between 2.5 and4.6 V; (b) Cyclic voltammograms of Li/Li(Ni1/3Co1/3Mn1/3)O2 cellbetween 2.5 and 4.8 V at a scan rate of 0.05 mV S�1 for the 1st, 2nd,5th, and 10th cycle at room temperature. Li metal acts as both counter andreference electrode.

Y. Liang et al. / Electrochemistry Communications 9 (2007) 965–970 967

formed by particles of 3–5 lm, with a modal size of 4.2 lm.Tap density data of three samples was also determined witha density of 2.41 g cm�3 for RP, 1.75 g cm�3 for RP and1.99 g cm�3 for CP, as displayed in Table 1. Although sub-micronic positive electrodes are often used to obtainincreased power, such an improvement is achieved at theexpense of volumetric energy density. The tap density ofspherical sample (RP) is close to commercial LiCoO2

(2.70 g cm�3).Fig. 3a shows the voltage vs. specific capacity profiles of

Li/Li(Ni1/3Co1/3Mn1/3)O2 cell (sample RP) between 2.5 and4.6 V at a constant rate of 0.5 C (100 mA g�1) for different

Table 1Tap density and specific energy density of three samples synthesized via differ

Sample Tap density (g cm�3) Gravimetric specifi

RP 2.41 203SP 1.75 167CP 1.99 204

cycles. The cell shows an initial discharge capacity of203 mAh g�1 with an irreversible capacity loss (ICL) of25 mAh g�1 during the first cycle and 188 mAh g�1 wasretained at the end of 100th cycle. On starting the current,the voltage ascended to 3.7 V, not similar to that observedby Ohzuku and Makimura [5], and then slowly increasedto 4.6 V. For the latter (as will be shown in Fig. 4), this phe-nomenon is caused by the polarization and hence an ineffi-cient charging due to low electronic conductivity of the

ent methods

c capacity (mAh g�1) Volumetric specific capacity (Ah L3)

489293406

Fig. 4. Typical initial charge/discharge profiles of three cells, synthesizedby RP (rheological phase reaction method), SP (solid state reactionmethod) and CP (co-precipitation method) between 2.5 and 4.6 V at 0.5 Crate.

Fig. 5. Comparison for pulse power ASI as a function of DOD for thegraphite/Li(Ni1/3Co1/3Mn1/3)O2 cells, synthesized by RP (rheologicalphase reaction method), SP (solid state reaction method) and CP (co-precipitation method): (a) 2-s charge; (b) 18-s discharge.

968 Y. Liang et al. / Electrochemistry Communications 9 (2007) 965–970

composite cathode prepared by traditional solid state reac-tion [4]. Moreover, the cell was found to show significantcapacity-fading when cycled between 2.5–4.6 V at 0.5 C.Cyclic voltammogram (CV) of the Li/Li(Ni1/3Co1/3-Mn1/3)O2 cell (sample RP) was carried out between 2.5and 4.8 V to study corresponding cathodic peaks and anodicpeaks position shift. The results are shown in Fig. 3b. Thefirst anodic scan shows a sharp and intense peak centredat 3.84 V and a second low-intensity peak at 4.65 V. The cor-responding first cathodic peaks are seen at 4.53 and 3.70 V,respectively. It can be seen that the first anodic peak getsshifted to lower voltage side by 0.12–0.14 V and the intensityreduced during next cycles, whereas the 3.81 V anodic peaksand 3.69 V cathodic peaks show no shift at all. Thus there isan effective decrease in the hysterisis between anodic andcathodic peaks in the 3.8 V region after the first cycle, indi-cating a better reversibility of the electrode [14].

In order to compare the effects on electrochemical prop-erties of the material caused by different methods, the ini-tial charge/discharge profiles of three Li/Li(Ni1/3-Co1/3Mn1/3)O2 cells were conducted at room temperature.The results of these cells cycled at a current density of0.5 C between 2.5 and 4.6 V are displayed in Fig. 4. SampleCP and RP deliver a charge capacity of 236, 228 mA g�1

and a discharge capacity of 204, 203 mAh g�1, respectively.A good initial efficiency of 89, 87% are presented respec-tively for above two samples, whereas SP gives a muchsmaller value of 78%. CP and RP show similar initialbehaviors at the first charging up to about 3.7 V, and thenslowly ascend up to 4.6 V. However, data of CP were sup-ported only in literatures [15,17,21,25] with the cycling per-formance more than 2 C rate due to the fast fade ofcapacity at high power.

To evaluate a possible application in HEV system,power pulse area-specific-impedance (ASI) as a functionof depth-of-discharge (DOD) for the graphite/Li-

(Ni1/3Co1/3Mn1/3)O2 cells is shown in Fig. 5, which are fab-ricated with three cathode materials prepared from differ-ent methods. After ten formation cycles at a slow currentrate (C/10), these cells were charged to a certain state ofcharge (SOC) and then disassembled in the glove box.Two symmetric cells (size 2016) were made out of each dis-assembled 15.5 cm2 electrode. Subsequent tests are accord-ing to the method reported in literature [24]. It is foundthat cell SP (solid state route) and CP (co-precipitationmethod) showed higher ASI than RP before DOD of90% regardless of the powder morphology. Besides, littlechange is found for cell RP during 2-s charge and 18-s dis-charge within DOD of 90%. Therefore, using Li(Ni1/3Co1/

3Mn1/3)O2 obtained from this newly developed method canreduce the impedance of a cell. The decrease of cell imped-ance as a result of sample RP favors the increase of cellcapacity. This can explain the little change of dischargecapacity as will be shown in Fig. 6.

Fig. 6 shows the capacity retention of Li/Li(Ni1/3-Co1/3Mn1/3)O2 cells (sample RP) at different discharge

Fig. 6. Rate performance of Li/Li(Ni1/3Co1/3Mn1/3)O2 cell by rheologicalphase reaction route at charge rate of 0.5 C and different discharge rate of0.5 C, 1 C, 2.5 C and 5 C, respectively.

Fig. 7. DSC traces for three Li(Ni1/3Co1/3Mn1/3)O2 electrodes charged to4.6 V at 0.1 C, synthesized by RP (rheological phase reaction method), SP(solid state reaction method) and CP (co-precipitation method) at aheating rate of 5 �C min�1.

Y. Liang et al. / Electrochemistry Communications 9 (2007) 965–970 969

rates (charge is at 0.5 C) in the range of 2.5–4.6 V. Thisregime was chosen to permit comparison with the resultson rate performance in the literatures [15,21,25,26]. Theinitial discharge capacity at different rates for the hydrox-ide and carbonate derived phases are lower, for example,153 mAh g�1 at 1.8 C (360 mA g�1) [26], 13 mAh g�1 at4 C [15], 125 mAh g�1 at 3.75 C [21] and 115 mAh g�1 at3.5 C [25]. The reversible capacity of the cell graduallydecreases in first nearly 10 cycles, then keep well stable.Spherical Li(Ni1/3Co1/3Mn1/3)O2 delivers a dischargecapacity of 188, 178, 159 and 147 mAh g�1 after 100 cyclesat a rate of 0.5 C, 1 C, 2.5 C and 5 C, respectively. Few sim-ilar results have been included at such current rate.Although high capacities with excellent capacity retentionmay be obtained at moderate rates (eg. 20 mA g�1), thefuture importance of high-power and high-energy applica-tions encourages investigation of the performance at highrate. Therefore, our Li(Ni1/3Co1/3Mn1/3)O2 material pro-vided improved reversible capacity and significant cyclingstability, which owns a capability for applications in EVand HEV systems.

Moreover, large-scale applications of lithium-ion batter-ies bring more rigid require for cathode with superiorsafety compared with small cells. It has been reported thatLi(Ni1/3Co1/3Mn1/3)O2 has superior safety characteristicsthan LiCoO2 [9]. DSC measurements of three Li(Ni1/

3Co1/3Mn1/3)O2 samples (RP, SP and CP) were carriedout to compare thermal stability of the electrodes. Eachapproximately 2.5 mg sample was loaded from electrodeswithout washing off the electrolyte. The total heat gener-ated by Li(Ni1/3Co1/3Mn1/3)O2 (CP) is much greater thanthat produced by SP and CP. The peak positions for thematerials derived from the solid state route and co-precip-itation method very similar to that observed previously[5,8]. In the case of this spherical material prepared usingrheological phase reaction method, the peak shifts to ahigher temperature of 319 �C implying somewhat improved

safety, compared to 302 �C and 313 �C for other two meth-ods, Fig. 7.

4. Conclusions

High tap-density Li(Ni1/3Co1/3Mn1/3)O2 microsphereswere prepared using a rheological phase reaction routesuitable for a large scale up. The electrochemical propertiesof the material have been much improved with reversiblecapacity and stable cycling performance at high rate, whichis capable of high power and high energy properties forapplications in EV and HEV systems.

Acknowledgement

This work was supported by the National Natural Sci-ence Foundation of China (20471044).

References

[1] J.N. Reimers, J.R. Dahn, J. Electrochem. Soc. 139 (1992) 209.[2] J.M. Kim, N. Kumagai, H.T. Chung, Electrochem. Solid State Lett. 9

(2006) A494.[3] C.H. Song, A.M. Stephan, S.K. Jeong SK, A.R. Kim, K.S. Nahm,

Y.S. Lee, J.K. Kim, H.T. Chung, Mater. Res. Bull. 41 (2006) 1487.[4] H. Gabrisch, J.D. Wilcox, M.M. Doeff, Electrochem. Solid State Lett.

9 (2006) A360.[5] T. Ohzuku, Y. Makimura, Chem. Lett. (2001) 642.[6] S. Patoux, M.M. Doeff, Electrochem. Commun. 6 (2004) 767.[7] B.J. Hwang, Y.W. Tsai, D. Carlier, G. Ceder, Chem. Mater. 15 (2003)

3676.[8] I. Belharouak, Y.K. Sun, J. Liu, K. Amine, J. Power Sources 123

(2003) 247.[9] Z. Lu, D.D. MacNeil, J.R. Dahn, Electrochem. Solid State Lett. 4

(2001) A200.[10] T. Nukuda, T. Inamasu, A. Fujii, D. Endo, H. Nakagawa, S.

Kozono, T. Iguchi, J. Kuratomi, K. Kohno, S. Izuchi, M. Oshitani, J.Power Sources 146 (2005) 611.

970 Y. Liang et al. / Electrochemistry Communications 9 (2007) 965–970

[11] Z. Wang, Y. Sun, L. Chen, X. Huang, J. Electrochem. Soc. 151 (2004)A914.

[12] J. Choi, A. Manthiram, Electrochem. Solid-State Lett. 7 (2004)A365.

[13] N. Yabuuchi, Y. Koyama, N. Nakayama, T. Ohzuku, J. Electro-chem. Soc. 152 (2005) A1434.

[14] K.M. Shaju, G.V. Subba Rao, B.V.R. Chowdari, Electrochim. Acta48 (2002) 145.

[15] J. Choi, A. Manthiram, J. Electrochem. Soc. 152 (2005) A1714.[16] T.H. Cho, S.M. Park, M. Yoshio, T. Hirai, Y. Hideshima, J. Power

Sources 142 (2005) 306.[17] S.H. Park, H. Shin, S.T. Myung, C.S. Yoon, K. Amine, Y.K. Sun,

Chem. Mater. 17 (2005) 6.[18] K.M. Shaju, P.G. Bruce, Adv. Mater. 18 (2006) 2330.

[19] M.H. Lee, Y.J. Kang, S.T. Myung, Y.K. Sun, Electrochim. Acta 50(2004) 939.

[20] N. Tran, L. Croguennec, C. Jordy, P. Biensan, C. Delmas, Solid StateIonics 176 (2005) 1539.

[21] G.H. Kim, J.H. Kim, S.T. Myung, C.S. Yoon, Y.K. Sun, Electro-chim. Acta 50 (2005) 4800.

[22] P. He, H.R. Wang, L. Qi, T. Osaka, J. Power Sources 160 (2006) 627.[23] Y. Makimara, T. Ohzuku, J. Power Sources 119–121 (2003) 171.[24] K. Amine, J. Liu, S. Kang, I. Belharouak, Y. Hyung, D.R. Vissers, G.

Henriksen, J. Power Sources 129 (2004) 19.[25] G.H. Kim, J.H. Kim, S.T. Myung, C.S. Yoon, Y.K. Sun, J.

Electrochem. Soc. 152 (2005) 1707.[26] T.H. Cho, S.M. Park, M. Yoshio, T. Hirai, Y. Hideshima, Chem.

Lett. (2004) 704.