Journal of Power Sources 109 (2002) 47–52
`
`Effect of Li2CO3 additive on gas generation in lithium-ion batteries
`Jee-Sun Shina, Chi-Hwan Hana, Un-Ho Jungb, Shung-Ik Leeb, Hyeong-Jin Kimc, Keon Kima,*
`aDivision of Chemistry and Molecular Engineering, Department of Chemistry, Korea University, Seoul 136-701, South Korea
`bDepartment of Chemical Engineering, Korea University, Seoul 136-701, South Korea
`cBattery Research Center, Research Park, LG Chemical Ltd., Taejon 305-380, South Korea
`
`Received 3 September 2001; accepted 10 January 2002
`
`Abstract
`
`To elucidate the mechanism of gas generation during charge–discharge cycling of a lithium-ion cell, the generated gases and passive films
`on the carbon electrode are examined by means of gas chromatography (GC) and Fourier transform infrared (FTIR) spectroscopy. In ethyl
`carbonate/dimethyl carbonate and ethyl carbonate/diethyl carbonate 1 M LiPF6 electrolytes, the detected gaseous products are CO2, CO, CH4,
`C2H4, C2H6, etc. The FTIR spectrum of the surface of the carbon electrode shows bands which correspond to Li2CO3, ROCO2Li,
`(ROCO2Li)2, and RCO2Li. These results suggest that gas evolution is caused by electrode decomposition, reactive trace impurities, and
`electrolyte reduction. The surface of the electrode is composed of electrolyte reduction products. When 0.05 M Li2CO3 is added as an
`electrolyte additive, the total volume of generated gases is reduced, and the discharge capacity and the conductivity of lithium-ions are
`increased. These results can be explained by a more compact and thin ‘solid electrolyte interface’ film on the carbon electrode formed by
`Li2CO3, which effectively prevents solvent co-intercalation and carbon exfoliation. # 2002 Elsevier Science B.V. All rights reserved.
`
`Keywords: Li-ion batteries; Additives; Gas generation; SEI; Li2CO3
`
`1. Introduction
`
`Lithium batteries are noted for their high specific energy
`compared with other
`secondary batteries. Recently,
`rechargeable lithium-ion batteries with capacities of
`1300–1900 mAh have been commercialized for some por-
`table electronic devices such as camcorders, computers, and
`cameras [1]. In addition, lithium-ion and lithium-polymer
`batteries are being developed as power sources for electric
`vehicles to provide longer driving ranges, higher accelera-
`tion, and longer lifetimes [2,3].
`Safety concerns have, however, limited the full utilization
`of lithium batteries [4–6]. Thus, battery safety is a key issue
`for the above applications. Electrolyte stability, in particular,
`is important for the entire safety of cells in practical use [7].
`Therefore, one of the important factors in the development
`of long-life and safe lithium-ion batteries is the estimation of
`the decomposition and compositional change of electrolytes
`and electrodes during charge–discharge cycling. In addition,
`the electrode for which active surface area is increased
`unnecessarily, results in an overall increase of reaction with
`electrolytes and organic solvents. The reactions can result in
`an irreversible loss of anode capacity, and gradual oxidation
`
`and consumption of electrolyte on the electrode. These
`effects, in turn cause capacity decline during cycling and
`an increased threat to battery safety by gas evolution [8].
`Many researchers have attempted to understand the
`mechanism of the degradation of the electrolyte on the
`carbon electrode, and have suggested the formation of films
`on the surface of the carbon electrode during initial charging
`[9–12]. Ein-Eli et al. [13–15] published a series of studies on
`the formation of passive films in lithium batteries. Also,
`there has been considerable research into improving elec-
`trolyte systems for lithium batteries by the use of small
`amount of additives [11,16–18].
`In this study,
`the electrolyte reduction and electrode
`decomposition are investigated by means of analysis of
`generated gases and passive films during repeated cycling
`of various electrodes and electrolyte systems, in order to
`elucidate the degradation mechanism of lithium-ion bat-
`teries. The main objective has been to reduce gas generation
`by using Li2CO3 as an electrolyte additive.
`
`2. Experimental
`
`2.1. Cells and electrode
`
`* Corresponding author. Tel.: þ82-2-953-1172; fax: þ82-2-953-1172.
`E-mail address: kkim@mail.korea.ac.kr (K. Kim).
`
`Carbon electrodes were prepared from MCMB 25–28
`material (25 mm, Osaka Gas Co.) by mixing the material
`
`0378-7753/02/$ – see front matter # 2002 Elsevier Science B.V. All rights reserved.
`PII: S 0 3 7 8 - 7 7 5 3 ( 0 2 ) 0 0 0 3 9 - 3
`
`APPLE-1014
`
`1
`
`

`

`48
`
`J.-S. Shin et al. / Journal of Power Sources 109 (2002) 47–52
`
`with a solution of poly (vinylidene fluoride) (PVDF, Aldrich;
`5 wt.%) dissolved in NMP (N-methyl 2-pyrrodinone,
`Aldrich). The slurry was then coated on a copper foil.
`The electrode was dried under vacuum at room temperature
`for 24 h.
`Cathodes were prepared by mixing LiCoO2 (10 mm,
`Seimi Co.) powder with carbon black, and a solution of
`PVdF dissolved in a weight ratio of 90:4:6. These compo-
`nents were well mixed using a mortar and pestle, and then
`were coated on aluminum foil. The electrode was dried
`under a vacuum at 120 8C for 24 h.
`Electrolytes comprised a 1 M solution of LiPF6 in a 1:1
`mixture (v/v) of ethylene carbonate (EC) and either diethyl
`carbonate (DEC) or dimethyl carbonate (DMC) (Merck,
`battery grade). A porous polypropylene film was used as
`the separator.
`
`2.2. Electrochemical measurements
`
`Charge–discharge measurements were performed with an
`Arbin cycler. Usually,
`the cells were cycled 50 times
`between 2.8 and 4.3 V at charging and discharging current
`densities of 0.5 mA cm2 (C/5 rate), and terminated in the
`discharged state at 2.8 V.
`The a.c. impedance measurements were carried out by
`means of an IM6 impedance analyzer (Zahner Electrik).
`Impedance spectra were potentiostatically measured by
`applying an a.c. voltage of 5 mV amplitude over the fre-
`quency range 100 MHz–100 kHz after the electrode had
`attained equilibrium at each potential. All the potentials
`indicated here are based on the Li/Liþ electrode potential.
`
`2.3. Instrumental measurements
`
`Fourier transform infrared spectroscopy (FTIR) measure-
`ments of carbon electrodes were obtained by using Bomem
`
`(MB-104) equipemnt. After cycling in a cell, the electrode
`was rinsed with dimethyl carbonate (DMC) to remove the
`salts, and then dried under vacuum at 100 8C for 24 h. The
`surface of the carbon electrode was scratched with a stain-
`less-steel knife. The powder was mixed with KBr and
`pelletized for IR measurement.
`The generated gas in the cell was analyzed by gas
`chromatography. A Hewlett-Packard 5890 series gas chro-
`matograph (GC), equipped with a flame ionization detector
`(FID), was used to detect hydrocarbons. A GOW-MAC GC,
`equipped with a thermal conductive detector (TCD), was
`used for composition analysis of the generated gases. For
`each sample, 0.5 ml (for TCD detector) and 5 ml (for FID
`detector) of gas were injected for these analyses. After cycle
`test, the gases generated in the cells were drawn into a gas-
`tight syringe through the septum.
`Experiments were performed with a two-electrode cell. In
`a lithium/carbon cell, the anode was lithium metal and the
`cathode was carbon. In a Li/LiCoO2 cell, the anode was
`lithium metal and the cathode was LiCoO2. In a lithium-ion
`cell, the anode was carbon and the cathode was LiCoO2.
`All experiments were conducted at room temperature and
`in ambient conditions. Cell assembly was conducted under a
`dry argon atmosphere in a glove box.
`
`3. Results and discussion
`
`3.1. Gas generation
`
`Tables 1 and 2 show the composition of generated gases
`during repeated cycling in ethylene carbonate/diethyl car-
`bonate (EC/DEC) and ethylene carbonate/dimethyl carbo-
`nate (EC/DMC), EC/DEC and EC/DMC 1 M LiPF6
`electrolytes. TCD was used to analyze the gas composition
`after 10, 30 and 50 cycles. The results show that the main
`
`Table 1
`Composition of generated gases during repeated cycles in EC/DEC 1 M LiPF6 electrolyte
`
`Generated gas
`
`Li/LiCoO2
`
`Li/C
`
`C/LiCoO2
`
`10 cycles
`
`30 cycles
`
`50 cycles
`
`10 cycles
`
`30 cycles
`
`50 cycles
`
`10 cycles
`
`30 cycles
`
`50 cycles
`
`CH4
`O2
`CO
`CO2
`
`18.004
`0.014
`23.68
`58.3
`
`20.378
`0.010
`21.99
`57.62
`
`21.057
`0.007
`24.00
`54.934
`
`7.254
`–
`21.467
`71.277
`
`9.963
`–
`21.66
`68.375
`
`13.267
`–
`23.74
`63.00
`
`7.11
`0.018
`21.63
`71.24
`
`13.131
`0.015
`26.31
`60.542
`
`15.108
`0.010
`29.01
`55.870
`
`Table 2
`Composition of generated gases during repeated cycles in EC/DMC 1 M LiPF6 electrolyte
`
`Generated gas
`
`Li/LiCoO2
`
`Li/C
`
`C/LiCoO2
`
`10 cycles
`
`30 cycles
`
`50 cycles
`
`10 cycles
`
`30 cycles
`
`50 cycles
`
`10 cycles
`
`30 cycles
`
`50 cycles
`
`CH4
`O2
`CO
`CO2
`
`27.094
`0.018
`18.469
`54.417
`
`26.153
`0.014
`15.657
`58.174
`
`25.191
`0.011
`13.625
`61.171
`
`22.008
`–
`36.153
`41.837
`
`28.819
`–
`35.005
`36.174
`
`29.111
`–
`32.334
`38.553
`
`15.721
`0.018
`27.774
`56.485
`
`23.191
`0.016
`22.126
`54.665
`
`25.171
`0.01
`17.638
`57.18
`
`2
`
`

`

`J.-S. Shin et al. / Journal of Power Sources 109 (2002) 47–52
`
`49
`
`gases are CO2, CH4, and CO. There is also a small amount of
`O2. Aurbach et al. [14,15] reported that EC and DMC
`undergo a reduction reaction to form lithium alkyl carbo-
`nates, CH4 and CO on lithium metal. The generation
`mechanisms of CH4 and CO are considered to be as follows
`[24,27]:
`
`DMC þ 2Liþ þ 2e þ H2 ! Li2CO3 # þ CH4 "
`
`2DMC þ 2Liþ þ 2e þ H2 ! CH3OLi # þ CH4 "
`
`DMC þ 2Liþ þ 2e ! CH3OLi # þ CO "
`
`EC þ 2Liþ þ 2e ! ðCH2OLiÞ # þ CO "
`
`DEC þ 2Liþ þ 2e ! CH3CH2OLi # þ CO "
`
`(1)
`
`(2)
`
`(3)
`
`(4)
`
`(5)
`
`As shown in Tables 1 and 2, CO and CH4 are mainly
`detected in the EC/DEC electrolyte and the EC/DMC elec-
`trolyte, respectively. These results are in good agreement
`with earlier studies which showed that the CH4 is generated
`mainly from DMC [14,15]. The data in Tables 1 and 2 also
`suggest that the main gaseous product is CO2, irrespective of
`the choice of electrolyte solvent. CO2 gas can be generated
`by two different processes. One is the decomposition of the
`cathode material which, during overcharging, is understood
`to be [28]:
`
`3CoO2 ! Co3O4 þ O2 "
`
`DMC þ 3O2 ! 3CO2 " þ H2O
`
`(6)
`
`(7)
`
`The other is the reaction via trace impurities (HF and
`H2O) in the electrolyte [29]:
`
`LiOCO2CH3 þ HF ! LiF þ CH3OH þ CO2 "
`
`(8)
`
`ðCH2OCO2LiÞ2 þ H2O ! Li2CO3 # þ 2CH2OH þ CO2 "
`(9)
`
`Eq. (6) shows that O2 is generated as a result of decom-
`position of the overcharged cathode material. While, Eq. (7)
`shows that CO2 is produced as a result of the subsequent
`reaction of oxygen with the electrolyte. When LiCoO2 is not
`used as a cathode, O2 is not detected. Kumari et al. [28] have
`also claimed that O2 generation is due to the degradation of
`the cathode material.
`Eq. (8) shows that CO2 is a final product on the lithium
`surface in electrolytes which contain HF as a contaminant
`(e.g. EC and DMC/DEC containing LiPF6) [16]. This result
`is essentially explained by acid–base reactions of HF with
`various basic lithium compounds in surface films. Also, the
`CO2 present may be formed by a secondary reaction of
`ROCO2Li with trace water. The solutions unavoidably con-
`tain 20–30 ppm of water, which is sufficient (taking into
`account the large volume-to-Li-surface-area ratio) to affect
`markedly the surface films in the time scale of the experi-
`ments. Therefore, the generation of a large amount of CO2 is
`mainly due to the decomposition of the cathode material and
`the reaction of trace impurities.
`
`Fig. 1. Composition of detected gases during repeated cycling in the EC/
`DEC 1 M LiPF6 electrolyte (DEC set as 100).
`
`The generated gases detected by FID during repeated
`cycling in EC/DEC and EC/DMC 1 M LiPF6 electrolytes are
`shown in Figs. 1 and 2, respectively. The value of y-axis
`represents the percentage composition of detected gases
`using DEC or DMC as an internal standard. As DEC or
`DMC gas produces its own gas, it is reasonable that DEC or
`DMC can be used as an internal standard. FID was used to
`detect generated hydrocarbon gases during the charge–dis-
`charge process. For the EC/DEC electrolyte, the gases are
`CH4, C2H4, C2H6, C3H6 and C3H8. C3H6, however, was not
`detected in the EC/DMC electrolyte. The generation
`mechanisms of the hydrocarbon gases, except for CH4,
`are considered to be as follows [23,24,26,27].
`
`DEC þ 2Liþ þ 2e þ H2 ! Li2CO3 # þ 2C2H6 "
`
`(10)
`
`DEC þ 2Liþ þ 2e þ H2 ! CH3CH2CO2OLi # þ C2H6 "
`(11)
`
`CH3OCO2CH3 þ e þ Liþ ! CH3OCO2Li # þ CH3
`
`
`
`CH3OCO2Li þ e þ Liþ ! Li2CO3 # þ CH3
`
`
`
`3
`
`

`

`50
`
`J.-S. Shin et al. / Journal of Power Sources 109 (2002) 47–52
`
`Fig. 2. Composition of generated gases during repeated cycling in EC/
`DMC 1 M LiPF6 electrolyte with/without Li2CO3 additive (DMC set as
`100).
`
`CH3
`
`CH3
`
` þ 1
`2 H2 ! CH4 "
` þ CH3
` ! C2H6 "
`
`(12)
`
`(13)
`
`C2H5OCO2C2H5 þ e þ Liþ ! CH3OCO2Li # þ C2H5
`(14)
`
`
`
`C2H5OCO2Li þ e þ Liþ ! Li2CO3 # þ C2H5
`
`
`
`evolution decreases during repeated cycling, irrespective of
`the electrode under study. This result can be explained by
`prior reduction of EC on the carbon surface [19]. It is
`thought that the cyclic structure of EC can contact more
`closely the surface of the electrode than the linear structure
`of DEC or DMC. In addition,
`the polarity and dipole
`moments of EC are greater than those of DEC and DMC
`[19]. Comparing the lithium metal anode and carbon anode,
`greater amounts of gases are generated with the lithium
`electrode due to the high reactivity of this metal.
`FTIR spectra obtained from carbon, taken from electrodes
`after 50 cycles in the salt solution (EC/DMC and EC/DEC
`1 M LiPF6), are presented in Fig. 3. The spectra analysis is
`based on previous work [23,24,26,27]. In the EC/DEC 1 M
`LiPF6 electrolyte, bands which correspond to Li2CO3
`(n ¼ 1510–1450 and 875–860 cm1), ROCO2Li, (ROCO2-
`Li)2 and RCO2Li (n ¼ 1640–1620, 1450–1400, 1350–1300
`and 1100 cm1) are observed. Similar results are also
`obtained for the EC/DMC electrolyte. On the basis of the
`spectral studies, the peaks are related to the reduction of
`electrolyte and lithium salts. This suggests that the passive
`film on the carbon electrode is formed by the reduction
`products of electrolyte, as shown in previous mechanisms.
`
`3.2. Additive effect
`
`According to the above mentioned gas generation
`mechanisms, Li2CO3 is one of the electrolyte reduction
`products. From the reversible reduction reaction,
`it
`is
`thought that the amount of gas decreases on adding Li2CO3
`in the electrolyte. As already noted [20], it is well known that
`Li2CO3 is produced as an electrochemical reduction product
`by bubbling CO2 in the electrolyte. Unfortunately, however,
`the solubility of CO2 in common lithium battery solvents is
`poor. Therefore, we directly added 0.05 M Li2CO3 to the
`EC/DMC electrolyte to examine the effect of this additive.
`The composition of the generated gases during repeated
`cycling in EC/DMC electrolyte and EC/DMC electrolyte
`containing 0.05 M Li2CO3 is shown in Fig. 2. There are two
`important changes on adding Li2CO3. One is the virtual
`disappearance of C2H6 gas. In Eq. (13), C2H6 is formed by
` radical from the reduction of DMC. This suggests
`the CH3
`that adding Li2CO3 to the electrolyte suppresses the forma-
` radicals. The other change on adding Li2CO3
`tion of CH3
`is a reduction in the total amount of generated hydrocarbon
`gases. For an EC/DMC electrolyte system which contains
`0.05 M Li2CO3, the total amount of generated hydrocarbon
`gases is 3.220% in a Li/LiCoO2 cell, and 1.184% in a C/
`LiCoO2 cell. Compared with the non-additive system, there
`is a decrease of 0.63% and 0.96% in the total amount
`of gases in the Li/LiCoO2 cell and the C/LiCoO2 cell,
`respectively.
`The discharge capacity as a function of cycle number in
`EC/DMC electrolyte and EC/DMC electrolyte containing
`0.05 M Li2CO3 is shown in Fig. 4. A higher degree of
`discharge capacity and a long cycle-life are obtained with
`
`C2H5
`
`C2H5
`
` þ 1
`2 H2 ! C2H6 "
` þ CH3
` ! C3H8 "
`In Eqs. (1), (2), (10) and (11), hydrocarbons, except for
`C2H4, are produced from the reduction of DMC and radical
`reactions. C2H4 is only generated by the reduction of EC, i.e.
`[17]
`EC þ 2Liþ þ 2e ! Li2CO3 # þ C2H4 "
`2EC þ 2Liþ þ 2e ! ðCH2OCO2LiÞ2 # þ C2H4 "
`The data in Figs. 1 and 2 show that C2H4 is the main
`gas product during the first cycle. While the total amount
`of detected gas continuously increases, the rate of C2H4
`
`(15)
`
`(16)
`
`(17)
`
`(18)
`
`4
`
`

`

`J.-S. Shin et al. / Journal of Power Sources 109 (2002) 47–52
`
`51
`
`Fig. 3. FTIR spectrum of carbon electrode after 50 cycles in: (a) Li | EC/DEC 1 M LiPF6 | carbon cell; (b) LiCoO2 | EC/DEC 1 M LiPF6 | carbon cell; (c)
`Li | EC/DMC 1 M LiPF6 | carbon cell; (d) LiCoO2 | EC/DMC 1 M LiPF6 | carbon cell.
`
`the electrolyte containing Li2CO3. The impedance spectrum
`for the first and tenth cycle is presented in Fig. 5. The first
`semi-circle, which corresponds to the interfacial impedance
`of the surface film formed on the carbon electrode, is
`observed [21–24]. The interfacial impedance for the EC/
`DMC electrolyte with Li2CO3 is smaller than that for the
`EC/DMC electrolyte alone.
`In general, when carbon is electrochemically charged
`with Liþ-ions in an electrolyte, the intercalation reaction
`is accompanied by an irreversible process in which solu-
`tion components are reduced on the carbon electrode.
`This behaviour has been said to form a ‘solid electrolyte
`interface’ (SEI) film [25] This prevents the co-intercalation
`
`of solvent molecules into the graphite, a process that may
`result in increased irreversible charge consumption and
`sometimes even in graphite exfoliation. The reduction pro-
`ducts include alkyl carbonate, lithium alkoxide, lithium
`alkyl carbonate, reduced anions, etc. Among them, it is well
`known that Li2CO3 is an effective agent for forming SEI
`films on carbon electrodes [26]. When Li2CO3 is used as an
`electrolyte additive, the passive film has already built up in
`Li2CO3 before the reduction reaction occurs. Therefore, it
`effectively suppresses solvent co-intercalation reactions and
`graphite exfoliation. This is supported by the observation
`
`Fig. 4. Discharge capacity as function of cycle number of carbon | LiCoO2
`(*) EC:DMC ¼ 1:1 1 M
`(&) EC:DMC ¼ 1:1 1 M LiPF6;
`cell
`LiPF6 þ 0:05 M Li2CO3.
`
`Fig. 5. Cole–Cole plots after first charge–discharge cycle of carbon | LiCoO2
`cell.
`
`5
`
`

`

`52
`
`J.-S. Shin et al. / Journal of Power Sources 109 (2002) 47–52
`
`that the interfacial impedance (Fig. 5) for the EC/DMC
`electrolyte with Li2CO3 is smaller than that for the EC/DMC
`electrolyte alone.
`
`4. Conclusions
`
`From the results of gas chromatographic analysis, the
`generated gaseous products during repeated cycling of
`lithium-ion cells are CO2, CO, O2, CH4, C2H4, C2H6,
`C3H6 and C3H8. The composition of the detected gases is
`influenced by the nature of the electrolyte and the electrode.
`The main gaseous product is CO2 and is produced mainly by
`the decomposition of the cathode material and the reaction
`of trace impurities. The hydrocarbons (CH4, C2H4, C2H6,
`etc.) and CO are produced by the reduction of DEC, DMC
`and EC. The generated gas compositions continuously
`increase during repeated cycling.
`By adding 0.5 M Li2CO2 as an additive, C2H6 is not
` radicals from the
`detected, because the formation of CH3
`reduction of DMC is suppressed, and the total amount of
`generated gases is decreased. Also, cell performance is
`enhanced. These results can be explained by the formation
`of a more conductive SEI on the surface of the carbon anode,
`as confirmed by impedance measurements. Thus, it can be
`concluded that Li2CO3 is an effective additive in reducing
`safety problems with lithium-ion batteries.
`
`Acknowledgements
`
`The authors gratefully acknowledge the direct support of
`this work by a Korea University Grant the Ministry of
`Education and Human Resources Development (BK21)
`and the LG Chemical Ltd.
`
`References
`
`[1] K. Nishjo, N. Furukawa, in: J.O. Besenhard (Ed.), Handbook of
`Battery Materials, vol. 1, Wiley, 1998 (chapter 2) pp. 47–50.
`
`[2] J.M. Tarascon, D. Guyomard, Electrochimica Acta 38 (1993)
`1221.
`[3] S. Megahed, B. Scrosati, Electrochem. Soc. Interface 3 (1994)
`34.
`[4] H. Maleki, G. Deng, A. Anani, J. Howard, J. Electrochem. Soc. 146
`(1999) 3224.
`[5] K. Xu, S.P. Ding, T.R. Jow, J. Electrochem. Soc. 146 (1999) 4172.
`[6] S.-I. Tobishima, J.-I. Yamaki, J. Power Sources 81 (1999) 882.
`[7] K.M. Abraham, D.H. Dasquariello, F.J. Maltin, J. Electrochem. Soc.
`133 (1986) 661.
`[8] Z.X. Shu, R.S. McMillan, J.J. Murray, I.J. Davidson, J. Electrochem.
`Soc. 140 (1993) 922.
`[9] R. Fong, U. von Sucken, J.R. Dahn, J. Electrochem. Soc. 137 (1990)
`2009.
`[10] D. Aurbach, Y. Ein-Ely, A. Zaban, J. Electrochem. Soc. 141 (1994)
`L1.
`[11] D. Aurbach, Y. Ein-Ely, O.Y. Chusid, Y. Carmeli, M. Babai, H.
`Yamin, J. Electrochem. Soc. 141 (1994) 603.
`[12] Y. Matsumura, S. Wang, J. Mondori, J. Electrochem. Soc. 142 (1995)
`2914.
`[13] Y. Ein-Eli, B. Markovsky, D. Aurbach, Y. Carmeli, Electrochim.
`Acta 39 (1994) 2559.
`[14] D. Aurbach, Y. Ein-Eli, B. Markovsky, A. Zaban, S. Luski, Y.
`Carmeli, H. Yamin, J. Electrochem. Soc. 142 (1995) 2882.
`[15] Y. Ein-Eli, D. Aurbach, J. Power Sources 54 (1995) 281.
`[16] R. McMillan, H. Slegr, Z.X. Shu, W. Wang, J. Power Sources 81/82
`(1999) 20.
`[17] J.O. Besenhard, M.W. Wagner, M. Winter, A.D. Jannakoudakis, P.D.
`Jannakoudakis, E. Theodoridou 43 (1993) 413.
`[18] A. Naji, J. Ghanbaja, P. Willmann, D. Billaud, Electrochim. Acta 45
`(2000) 1893.
`[19] C.R. Yang, Y.Y. Wang, C.C. Wan, J. Power Sources 72 (1998) 66.
`[20] P.A. Christensem, A. Hamnett, A.V.G. Muir, N.A. Freeman, J.
`Electroanal. Chem. 288 (1990) 197.
`[21] M.D. Levi, D. Aurbach, J. Phys. Chem. B 101 (1997) 4630.
`[22] A. Funabiki, M. Inaba, Z. Ogumi, S. Yuasa, J. Otsuji, A. Tasaka, J.
`Electrochem. Soc. 145 (1998) 172.
`[23] D. Aurbach, A. Zaban, J. Electrochem. Soc. 141 (1994) 1808.
`[24] D. Aurbach, A. Zaban, Y. Ein-Eli, I. Weissman, O. Chusid, B.
`Markovsky, M. Levi, E. Levi, A. Schechter, E. Garnot, J. Power
`Sources 68 (1997) 91.
`[25] Y.C. Chang, H.J. Sohn, J. Electrochem. Soc. 147 (2000) 50.
`[26] Y. Ein-Eli, S.R. Thomas, V.R. Koch, J. Electrochem. Soc. 144 (1997)
`1159.
`[27] H. Yoshida, T. Fukunaga, T. Hazama, M. Terasaki, M. Mizutani, M.
`Yamachi, J. Power Sources 68 (1997) 311.
`[28] K. Kumai, H. Miyashiro, Y. Kobayashi, K. Takei, R. Ishikawa, J.
`Power Sources 81 (1999) 715.
`[29] D. Aurbach, A. Zaban, J. Electrochem. Soc. 142 (1995) L108.
`
`6
`
`