Prospective cathode materials LiNi0.7–xMgxCo0.3O2 (0≤x≤0.1) for a lithium-ion secondary battery were synthesized using a sol-gel method. The structural and electrochemical properties were examined by means of X-ray diffraction (XRD), scanning electron microscopy (SEM), cyclic voltammetry(CV), and charge-discharge tests. The results show that the LiNi0.7–xMgxCo0.3O2 maintains the α-NaFeO2 type layered structure regardless of the magnesium content in the range x⩽0.1. On the other hand, Mg-doping improves the capacity retention well. Besides, the Mg-doping promotes the diffusion of Li+ in LiNi0.7Co0.3O2. Moreover, Mg-doping suppresses the phase transitions that usually occur in LiNiO2 during cycling and improves the charge-discharge reversibility of Li/LiNi0.7Co0.3O2. High temperature cycling performance of the cathode at 55.5°C is also improved by Mg-doping, which is possibly attributed to the total stronger metal-oxygen bonding and the enhanced structure stability of those delithiated Mg-doped cathodes during cycling.
1. Introduction
Research in cathode materials for rechargeable lithium-ion batteries has a great deal of interest on the layered LiNiO2 and LiCoO2 materials in recent years [1–3]. Doped LiNiO2 may possess better electrochemical properties than LiNiO2. Lithium nickel cobalt oxide [4–6] is becoming one of the most promising positive electrode materials for lithium-ion batteries. Lithium nickel cobalt oxide (LiNiCoO2) for practical lithium-ion polymer battery production has become commercially available from several Japanese companies [7]. Recently, much attention has been focused on the development of LiNi1–yCoyO2 (0≤y≤1) systems, which have strong potential for commercial application [8]. Of the various y values, y=0.2 to 0.3 has produced the most interesting electrochemical properties [9]. LiNi0.74Co0.26O2 has been synthesized via a solution route with excellent electrochemical properties [10] and Mg-doped LiNi0.74Co0.26O2 with excellent electrochemical properties has been synthesized by a particular sol-gel method [11]. The reversible capacity of LiNi0.7Co0.3O2 tends to decrease during cycling and the thermal stability needs to be improved. Lattices doped by Al [12, 13], Mg [14], Mn [15, 16], Mg/Ti [17], and so on were investigated to solve the problems. In fact, Mg2+ has no 3d state, yet the 3d state of Mn4+ is 3d3(t2g3e2g0), different from Cr6+3d0(t2g0e2g0).
It was reported that Fe and Mg substitute for Ni in LiNi0.7Co0.2Ti0.05Mg0.05O2 [18] and LiNi0.6–xMgxCo0.25Mn0.15O2 [19] could reduce the cation mixing, improving structural integrity and cycle stability. Obviously, substitute for Ni in the Ni-based oxide appeares to be a good method to modify the structural and electrochemical performance of these materials. In this study, we employed magnesium (Mg2+) as a dopant due to the more negative Gibbs free energies of the MgO [ΔfGMgOθ=-610kJmol-1] compared to those of NiO [ΔfGNiOθ=-211.7kJmol-1]. Therefore, by substituting a part of Ni with Mg in layered LiNi0.7Co0.3O2 the total metal-oxygen bonding of the doped material is stronger than that of undoped one. Hence, I have synthesized a novel layered material of LiNi0.7–xMgxCo0.3O2 by substituting a small amount of Ni with Mg. The structures and electrochemical properties of the compounds have been investigated in this paper.
2. Experimental
LiNO3(AR), Ni(NO3)2·6H2O (AR), Co(Ac)2·4H2O(AR), and Mg(NO3)2·6H2O (AR) were used as starting materials of lithium, nickel, cobalt, and magnesium in LiNi0.7–xMxCo0.3O2 (0⩽x⩽0.1), respectively. LiNO3 was initially dissolved in citric acid solution. The amount of citric acid is equal to the total molar amount of Co, Ni, and Mg. Then, Co(Ac)2·4H2O, Ni(NO3)2·6H2O and Mg(NO3)2·6H2O were added to the mixture. The whole mixture was heated by water bath at 80°C. During the heating process, a clear, pink solution without any precipitation formed. At last, the clear solution was slowly dried and turned into gel. The xerogel was dried, ground, and then heat-treated in an oven at 120°C for 12 h. The gel precursor was calcined at 500°C in air for 6 h, and cooled to room temperature in a tube-furnace. The heat-treated products were ground in an agate mortar to obtain powders. And then the powder was calcined at 800°C for 12 h.
The phase identity and crystal structure of the materials were investigated by measuring X-ray diffraction (XRD) using a D-8 X-ray diffractometer (Bruker, Germany). The surface morphology was observed by means of a scanning electron microscopy (SEM, SIRION, FEI) with Cu Kα radiation at 40 kV, 70 mA.
For fabrication of cathodes, the prepared products were first mixed with acetylene black and polyvinylidene fluoride (80 : 8 : 12 in weight) in N-methylpyrrolidone(NMP). The slurry obtained was then coated onto Al foil and dried at 80°C for 18 h for further roll pressing. The electrochemical tests were done using CR2032 coin-type cells, which consist of the cathode, lithium foil as the anode, cegard 2325 as the separator, and 1 mol dm−3 LiPF6 in ethylene carbonate (EC)-diethyl carbonate (DEC) (1 : 1 in volume, LB302 from Guotai Huarong, China). The active material of cathode is 10.5 mg for each cell. The cells were charged and discharged at room temperature or 55°C at 0.1 C in the voltage range of 2.0–4.4 or 4.6 V (versus Li/Li+).
Cyclic voltammograms of LiNi0.7–xMgxCo0.3O2 (0⩽x⩽0.1) were obtained by IM6 in the voltage range 2.0–4.8 V at a scanning rate of 0.1 mVs−1. The reference and counter electrodes were fabricated from lithium metal.
3. Result and Discussion
Figure 1 shows the XRD spectra of LiNi0.7–xMgxCo0.3O2 (0⩽x⩽0.1) powders. All peaks can be indexed based on a hexagonal α-NaFeO2 structure with a space group of R3-m. No impurity-related peaks are observed from the XRD patterns with Mg-doping. The peaks are sharp and well defined, suggesting that the compounds are well crystallized. The clear splitting of the peaks 006/012 and 108/110 indicates an ordered distribution of cations in the layered structure [1, 20]. The lattice constants, a and c, c/a, and I(003)/I(104) of the LiNi0.7–xMgxCo0.3O2 (0⩽x⩽0.1) are shown in Table 1. From Table 1, one can see that with the Mg content increasing, the lattice constants a increase from 2.8547 to 2.8601 and cincreases and then decreases. Larger ionic size of Mg2+ (0.660 Å) than Ni3+ (0.560 Å) might cause the increase in lattice parameters. The unit cell volume expansion caused by the increase of lattice parameters may assist the intercalation and deintercalation of Li ions during electrochemical processes. Meanwhile, c/a ratio is an indication of the cation ordering in the layered cathode materials [21]. For the synthesized LiNi0.7–xMgxCo0.3O2 (0⩽x⩽0.1), all the c/a ratios are found to be higher than 4.9, indicating the higher cation ordering in all samples [21]. Besides, the ratio of I(003)/I(104) is also used as an indicator for cation mixing in the structure of layered cathode materials [17]. The Mg-doped materials also present a little higher I(003)/I(104) values than the undoped material. Similar to the change of c/a ratio, the increase for I(003)/I(104) is not linear with the increase of Mg content and LiNi0.65Mg0.05Co0.3O2 shows the highest I(003)/I(104) ratio. These results indicate that Mg-doping improves the characteristics of layer structure for LiNi0.7Co0.3O2 and LiNi0.65Mg0.05Co0.3O2 has the best layered characteristic.
Results of lattice parameters by XRD analysis of LiNi0.7-xMgxCo0.3O2 (0≤x≤0.1).
Sample
a(Å)
c(Å)
c/a
I(003)/I(104)
V(Å3)
LiNi0.7Co0.3O2
2.8547
14.1360
4.952
1.313
99.76
LiNi0.65Mg0.05Co0.3O2
2.8593
14.1724
4.957
1.541
100.35
LiNi0.6Mg0.1Co0.3O2
2.8601
14.1594
4.951
1.486
100.31
XRD patterns of LiNi0.7–xMgxCo0.3O2: (a) x=0; (b) x=0.05; (c) x=0.1.
Scanning electron micrograph was performed to characterize the approximative grain sizes and surface morphologies of the LiNi0.7–xMgxCo0.3O2 (0≤x≤0.1). As shown in Figure 2, the undoped material appears to be homogeneous size distribution with submicron particle size. In fact, the particle feature of this material synthesized by sol-gel method is favourable for the intercalation-deintercalation reaction of electrode during the charge-discharge processes, and it is expected to deliver a larger capacity. A similar size distribution (about 240 nm) and morphology is observed for Mg-doped materials regardless of the Mg contents. This suggests that the Mg is well permeated into the bare LiNi0.7Co0.3O2 to form a solid solution, indicating that Mg ions are homogeneously dissolved into the colloidal precursor in the solution prior to following heat treatments for calcination. Besides, the smaller particle size of LiNi0.7–xMgxCo0.3O2 tends to improve the electrochemical performance of the electrode by reducing the ion diffusion pathway during Li+ intercalation and deintercalation processes [11].
Scanning electron micrograph obtained for LiNi0.7–xMgxCo0.3O2 (0≤x≤0.1): (a) x=0.00; (b) x=0.05; (c) x=0.1.
The electrochemical performances of the studied compounds have been evaluated at room temperature and high temperature, respectively. The charge-discharge curves of the studied phase appear to be quite monotonous indicating that no major structural transitions occur during the lithium extraction/insertion reactions.
Figure 3 presents the voltage versus capacity profiles of all doped and undoped LiNi0.7Co0.3O2 materials in the first charge/discharge cycle. During the first cycle, the cell efficiency was 91.3% for x=0. The first coulomic efficiencies decrease with gradual substitution of Mg with Ni up to certain limit (x=0.1).
Initial charge-discharge curves of LiNi0.7–xMgxCo0.3O2 (0≤x≤0.1) at 0.1 C in 2.0–4.4 V: (a) x=0.05; (b) x=0; (c) x=0.1.
Charge-discharge characteristics of LiNi0.7–xMgxCo0.3O2 (0≤x≤0.1) were investigated by performing cycle tests in the range of 2.0–4.4 V. The initial discharge capacities for LiNi0.7–xMgxCo0.3O2 (0≤x≤0.1) are 178.9 (x=0.05), 171.7 (x=0), and 147.5 mAh g−1 (x=0.1), respectively. The increase in specific capacity for LiNi0.65Mg0.05Co0.3O2 could be attributed to the enhancement of layered characteristics with Mg-doping. For LiNiO2 and LiNi1–xCoxO2 cathodes, a capacity loss always occurs at the initial charge-discharge cycle due to the cation mixing [22]. The sample LiNi0.7–xMgxCo0.3O2 has less cation mixing so it has the less capacity loss. Therefore, it is easy to understand that LiNi0.65Mg0.05Co0.3O2 has the smallest capacity loss.
Figure 4 shows the charge-discharge curves versus cycle number during the first and the 20th cycles for LiNi0.7Co0.3O2 at C/10 rate in the 2.0–4.4 V range. The reversible capacity obtained for the first cycle is rather good (>150 mAh/g) with an irreversible capacity loss of 2.1 mAh/g. As shown in the derivative curve, the Li extraction occurs at 3.7 V which is equal to that obtained in the case of LiNiO2 [23]. This result confirms that previously obtained result by Han et al. [24] that Ni3+ is preferentially oxidized to Ni4+ compared to Co3+.
Charge-discharge curves versus cycle number (1st and 20th) of LiNi0.7Co0.3O2 at C/10 rate in the 2.0–4.4 V range.
To understand the effect of Mg-doping on the cycling stability of LiNi0.7Co0.3O2, the cycling performance of LiNi0.7–xMgxCo0.3O2 (0≤x≤0.1) in the voltage 2.0–4.6 V at the 0.1 C charge-discharge rate was investigated. The curves of discharge capacity versus cycle number for Li/LiNi0.7–xMgxCo0.3O2 (0≤x≤0.1) cell are shown in Figure 5. From Figure 5, it can be seen that the discharge capacity of bare LiNi0.7Co0.3O2 decreases to 158.1 mAh g−1 after 20 cycles with capacity retention of 87%. The cycling performance of Mg-doped materials are improved. For x=0.05, the capacity retention after 20 cycles is 98.4%. Evidently, the cycling reversibility of LiNi0.7Co0.3O2 cathode is improved significantly by Mg-doping. The better capacity retention of Mg-doped LiNi0.7Co0.3O2 during cycling tests could be attributed to that the doped Mg assists maintenance of the original layered crystal structure during intercalation-deintercalation of Li ions. In other words, the magnesium ions, with a size close to that of lithium, remain in the divalent state during cell charge-discharge. Therefore, their presence in the interslab space does not strongly affect lithium reintercalation because no shrinkage of the structure occurs upon cycling. This result explains how the magnesium-substituted phases have good cycling properties with enhanced capacity. The presence of an optimum concentration of magnesium ions in the lithium site prevents any local collapse of the interslab space during the deintercalation process. At a higher dosage of Mg2+ ions in the cathode material, the Ni2+ preferentially occupies the interslab spacing during the electrochemical process, which causes the low capacity delivery. On the other hand, the presence of Mg2+ ions in the lithium sites at the optimum concentration significantly reduces the usual changes observed in the cell parameters and accounts for good capacity retention [25].
Discharge capacity versus cycle number for Li/LiNi0.7–xMgxCo0.3O2 (0≤x≤0.1) cell at 0.1 C in 2.0–4.6 V: (a) x=0; (b) x=0.05; (c) x=0.1.
The cycling discharge capacities for LiNi0.65Mg0.05Co0.3O2 in the different voltage regions are presented in Figure 6. The excellent cycling stabilities in both voltage regions show that small particle size and good surface morphology could be considered as good features to obtain this excellent electrochemical performances.
Discharge capacities for a Li/LiNi0.65Mg0.05Co0.3O2 cell during the first 40 cycles at C/10 rate in the different voltage ranges.
Figure 7 shows the discharge capacity curves for Li/LiNi0.7–xMgxCo0.3O2 (0≤x≤0.1) cell during the first 20 cycles at C/10 rate and at 55°C in the 2.0–4.4 V range. Thermal stability of the charged cathodes at 55°C is also improved by Mg-doping, which is possibly attributed to the the total stronger metal-oxygen bonding and the enhanced structure stability of those delithiated Mg-doped cathodes during cycling.
Discharge capacity versus cycle number for Li/LiNi0.7–xMgxCo0.3O2 (0≤x≤0.1) cell at 0.1 C and 55.5°C in 2.0–4.4 V: (a) x=0; (b) x=0.1.
In order to investigate the effect of extra Ni2+ ions in the lithium site on the rate capability of this material, the cells were cycled at different current rates 0.1 C, 0.2 C, 0.3 C, 1 C, and 2 C between 2.0 and 4.6 V. As shown in Figure 8, it appears clearly that this LiNi0.65Mg0.05Co0.3O2 material exhibits a good reversibility at different rates. Although the capacity decreases with increasing the rate, LiNi0.65Mg0.05Co0.3O2 could keep more than 96% of its capacity during 40 cycles for all the tested rates, even if for 1 C or 2 C higher rates. This confirms that lithium ion’s diffusion in this material becomes easier.
The specific discharge capacities versus cycle number of Li/LiNi0.65Mg0.05Co0.3O2 cells at different rate 0.1 C, 0.2 C, 0.3 C, 1 C, and 2 C in the voltage range 2.0–4.6 V ((a)=0.1C; (b)=0.2C; (c)=0.3C; (d)=1C; (e)=2C).
Cyclic voltammetry (CV) was used to investigate the electrochemistry of the cathode materials LiNi0.7–xMgxCo0.3O2, as a function of Mg contents. The shape of the CV curve, peak potentials, and peak currents represents the electrochemical properties of the electrode and discloses the phase transitions that occur during charge-discharge experiments, which strongly affect the capacity fading during cycle [26]. In general, when a cathode experiences phase transformation, a peak occurs in the CV curve due to the coexistence of two phases. During charge-discharge experiments, LiNiO2 shows four different phases (one monoclinic phase, M, and three hexagonal phases, H1, H2, and H3) with three peaks in the CV curve. The three peaks in LiNiO2 correspond to the coexistence of H1 and M, M and H2, and H2 and H3, respectively. LiCoO2 also shows four phases (one monoclinic phase, M, and three hexagonal phases, H1, H2, and H3) in the CV curve [27, 28]. The three peaks in LiCoO2 correspond to the coexistence of the phases H1 and H2, H2 and M, and M and H3.
In this research, cyclic voltammograms of bare LiNi0.7Co0.3O2 and Mg-doped LiNi0.7–xMgxCo0.3O2 are compared in order to deduce the cause of the better capacity retention of LiNi0.65Mg0.05Co0.3O2. The CV measured CV curves are presented in Figure 9. The curve clearly shows the presence of three peaks at different voltage positions for LiNi0.7Co0.3O2, which arise from multiple phase transitions during charge-discharge cycling. The second peak in the CV curve for LiNi0.7Co0.3O2, while diminishing in intensity, merges with the first peak with increasing Mg content. The oxidation peak at 3.65 V almost does not change. Mg-doping suppresses the phase transitions that usually occur in LiNiO2 during cycling and improves the charge-discharge reversibility of Li/LiNi0.7Co0.3O2. The simplified CV curves and the peak shift are associated with the suppression of phase transitions due to superior maintenance of the layered structure after Mg addition.
Cyclic voltammograms for LiNi0.7–xMgxCo0.3O2 (0≤x≤0.1) powders for x=0.00 (a), 0.05 (b), and 0.10 (c).
4. Conclusion
LiNi0.7–xMgxCo0.3O2 cathodes (x=0, 0.05, 0.1) with submicron particles have been successfully synthesized by a sol-gel method calcinated at 800°C for 12 h. The structural and electrochemical properties have been systemically investigated to examine the effects of Mg-doping on initial discharge capacity and capacity retention. The results show that all the prepared LiNi0.7–xMgxCo0.3O2 materials maintain the α-NaFeO2 type layered structure regardless of the magnesium content in the range x⩽0.1. Mg-doping improves the capacity retention significantly. Besides, the Mg-doping promotes the diffusion of Li+ in LiNi0.7Co0.3O2. Moreover, Mg-doping suppresses the phase transitions that usually occur in LiNiO2 during cycling and improves the charge-discharge reversibility of Li/LiNi0.7Co0.3O2. Thermal stability of the charged cathodes is also improved by Mg-doping, which is possibly attributed to the lowered oxidation ability and the enhanced structure stability of those delithiated Mg-doped cathodes during cycling. Furthermore, charge-discharge cycling at different rates reveals that Mg-doped material has rather excellent capacity retention and rate capability.
Conflict of Interests
The author declares that there is no conflict of interests regarding the publication of this paper.
YamakiJ.MakideraM.KawamuraT.EgashiraM.OkadaS.Voltage prediction of nano-sized LiNiO2 cathode for use in Li-ion cells200615322452502-s2.0-3334447150310.1016/j.jpowsour.2005.05.024KwonS. N.SongJ.MummD. R.Effects of cathode fabrication conditions and cycling on the electrochemical performance of LiNiO2 synthesized by combustion and calcination2011375154315482-s2.0-7995547750710.1016/j.ceramint.2011.01.028LiL.ChenR.SunF.WuF.LiuJ.Preparation of LiCoO2 films from spent lithium-ion batteries by a combined recycling process20111083-42202252-s2.0-7995797752510.1016/j.hydromet.2011.04.013LiD.PengZ.RenH.GuoW.ZhouY.Synthesis and characterization of LiNi1-xCoxO2 for lithium batteries by a novel method200810711711762-s2.0-3634899805110.1016/j.matchemphys.2007.06.069MukaiK.SugiyamaJ.AokiY.Structural, magnetic, and electrochemical studies on lithium insertion materials LiNi1-xCoxO2 with 0≤x≤0.2520101837172617322-s2.0-7795391673010.1016/j.jssc.2010.05.019FeyG. T.-K.ShiuR. F.KumarT. P.ChenC. L.Preparation and characterization of lithium nickel cobalt oxide powders via a wet chemistry processing200310032342432-s2.0-004267432710.1016/S0921-5107(03)00109-0http://www.seimichemical.co.jp, http://www.smm.co.jpLiD.PengZ.GuoW.ZhouY.Synthesis and characterization of LiNi0.9Co0.1O2 for lithium batteries by a novel method20084571-2L1L52-s2.0-4264909933810.1016/j.jallcom.2007.02.146ChangC.-C.KimJ. Y.KumtaP. N.Divalent cation incorporated Li(1+x)MMgxO2(1+x) (M = Ni0.75Co0.25): viable cathode materials for rechargeable lithium-ion batteries200089156632-s2.0-003373829910.1016/S0378-7753(00)00393-1ChoJ.ParkB.Preparation and electrochemical/thermal properties of LiNi0.74Co0.26O2 cathode material2001921-235392-s2.0-003519247610.1016/S0378-7753(00)00499-7ChoJ.LiNi0.74Co0.26-xMgxO2 cathode material for a Li-in cell20001230893094DuvigneaudP. H.SegatoT.Synthesis and characterisation of LiNi1-x-yCoxAlyO2 cathodes for lithium-ion batteries by the PVA precursor method2004246137513802-s2.0-094230188310.1016/S0955-2219(03)00565-XHanC. J.YoonJ. H.ChoW. I.JangH.Electrochemical properties of LiNi0.8Co0.2-xAlxO2 prepared by a sol-gel method200413611321382-s2.0-434466264810.1016/j.jpowsour.2004.05.006PouillerieC.PertonF.BiensanP.PérèsJ. P.BrousselyM.DelmasC.Effect of magnesium substitution on the cycling behavior of lithium nickel cobalt oxide20019622933022-s2.0-003587731210.1016/S0378-7753(00)00653-4MacNeilD. D.LuZ.DahnaJ. R.Structure and electrochemistry of Li[NixCo1-2xMnx]O2 (0≤x≤1/2)200214910A1332A13362-s2.0-003680422310.1149/1.1505633SunY.OuyangC.WangZ.HuangX.ChenL.Effect of Co content on rate performance of LiMn0.5-xCo2xNi0.5-xO2 cathode materials for lithium-ion batteries20041514A504A5082-s2.0-204250969210.1149/1.1647574GaoY.YakovlevaM. V.EbnerW. B.Novel LiNi1-xTix/2Mgx/2O2 compounds as cathode materials for safer lithium-ion batteries1998131171192-s2.0-0032164657SubramanianV.FeyG. T.-K.Preparation and characterization of LiNi0.7Co0.2Ti0.05M0.05O2 (M=Mg, Al and Zn) systems as cathode materials for lithium batteries20021483-43513582-s2.0-003660415810.1016/S0167-2738(02)00073-5LiaoP.-Y.DuhJ.-G.SheuH.-S.Structural and thermal properties of LiNi0.6-xMgxCo0.25Mn0.15O2 cathode materials200818327667702-s2.0-4824915273310.1016/j.jpowsour.2008.05.058GuaY. J.ChenY. B.LiuH. Q.Structural characterization of layered LiNi0.85-xMnxCo0.15O2 with x = 0, 0.1, 0.2 and 0.4 oxide electrodes for Li batteries20115097915792110.1016/j.jallcom.2011.05.024OhzukuT.NakuraK.AokiT.Comparative study of solid-state redox reactions of LiCo1/4Ni3/4O2 and LiAl1/4Ni3/4O2 for lithium-ion batteries1999451-21511602-s2.0-0033320452ChenC. H.LiuJ.AmineK.Symmetric cell approach and impedance spectroscopy of high power lithium-ion batteries20019623213282-s2.0-003587735910.1016/S0378-7753(00)00666-2GoverR. K. B.KannoR.MitchellB. J.HiranoA.KawamotoY.The role of nickel content on the structure and electrochemical properties of Lix(NiyCo1-y)O2200197-983163202-s2.0-003539744010.1016/S0378-7753(01)00623-1HanC. J.EomW. S.LeeS. M.ChoW. I.JangH.Study of the electrochemical properties of Ga-doped LiNi0.8Co0.2O2 synthesized by a sol-gel method200514412142192-s2.0-1884436989910.1016/j.jpowsour.2004.12.021SathiyamoorthiR.ShakkthivelP.RamalakshmiS.ShulY.-G.Influence of Mg doping on the performance of LiNiO2 matrix ceramic nanoparticles in high-voltage lithium-ion cells200717129229272-s2.0-3454838423610.1016/j.jpowsour.2007.06.023DokkoK.NishizawaM.HorikoshiS.ItohT.MohamediM.UchidaI.In situ observation of LiNiO2 single-particle fracture during Li-Ion extraction and insertion2000331251272-s2.0-003415872410.1149/1.1390977ReimersJ. N.DahnJ. R.Electrochemical and in situ x-ray diffraction studies of lithium intercalation in LixCoO219921398209120972-s2.0-0026908949OhzukuT.UedaA.Solid-state redox reactions of LiCoO2 (R̄3m) for 4 volt secondary lithium cells199414111297229772-s2.0-0028550468