Marys Medicine

Pii: s0927-796x(01)00030-4

M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 2. Principle of lithium ion battery.
A lithium ion rechargeable battery is also known as a swing battery or rocking chair battery since two-way movement of lithium ions between anode and cathode through the electrolyte occurs during charge and discharge processes. Increasing high voltages are associated with increasing difficulty in removing a lithium ion from a particular site during charge or discharge; the site with the lowest potential energy will be occupied first by lithium ions. The redox potential of the lithium- intercalated graphite anode does not differ much from the lithium anode potential and falls by only 0.2 V belowthat of lithium metal (i.e. the interaction between a lithium ion and the host carbon structure is small). Since lithium is intercalated into and not deposited over the surface of the graphite anode, the anode in a lithium ion battery differs from a pure lithium metal anode. Unlike the lithium metal anode, the graphite anode does not lend to dendrite formation during charging, and this eliminates any short circuits and thereby, provides safety to the cells.
A cylindrical-type battery is shown in Fig. 3 [2]. The aluminum foil, which is easy to process and is resistant to an oxidizing potential is used as the current collector for the cathode. Copper foil is used as the anode current collector since it does not react with lithium ion at low electrical potential.
Fine, porous, thin films of polyethylene or polypropylene are used as separators, and they operate orthopedically on a sheet with binders (Teflon, PVdF, etc.) and conductive additives (acetylene black, ketjen black, etc.), which is rolled in the shape of a cylindrical pipe. The ionic conductivity of the organic liquid electrolyte containing lithium ions is 10 2 S/cm at room temperature, which is too small compared to the conductivity of aqueous electrolytes. In order to minimize the internal resistance of the cell so as to secure sufficient current and energy capacity, the separator films are processed into thin sheets with large surface area. The materials selected for use should have excellent thermal stability. When the temperature in the battery rises during unwarranted conditions such as an external short circuit, decomposition of the electrolyte occurs resulting in a build-up of M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 3. A structure of cylindrical lithium ion battery.
gas pressure. To provide safety, a PTC element and a small gas discharge vent are placed inside. The lithium ion rechargeable battery marketed nowby the Sony Company (SONY US18650; height 65 mm, diameter 18 mm and weight 45 g) is high-tech type with a cut-off voltage control circuit charger and is reported to have a nominal capacity of 1800 mAh with cycle life as high as 500.
3. Materials development for lithium ion rechargeable battery 3.1. Anode materials In the lithium ion rechargeable battery marketed now, carbon is used as the anode-active material. Graphite is built of ABAB layers, which are held together by van der Waals forces; each layer contains conjugated sp2 bond. If lithium ions intercalate into these layers, the layer arrangement changes to AAA2 . Thus, graphite serves as the host structure for lithium intercalation and the structure is resilient enough to provide reversibility by allowing easy insertion and deinsertion of lithium. Graphite exists in various forms ranging from a crystalline state to an almost amorphous state, and the rawmaterial used to produce anodes include natural graphite, oil pitch, coal tar, hydrocarbon gas, benzene, various resins, etc. These anodes are categorized into (1) graphite, (2) non-graphitized glass-like carbon (hard carbon) which does not graphitize even if it is heat treated at high temperature and (3) soft carbon, easily changeable with heat treatment. The schematic profile of these carbons is shown in Fig. 4 [1]. Generally, graphite and hard carbon are used as anodes. In practice, at room temperature, graphite accepts sufficient lithium to form LiC6 which on delithiation can deliver 372 mAh/g. The intercalation of lithium into graphite and the electrochemical characteristics of graphite anodes have been reported by many researchers [3±5]. The relation between the electrochemical intercalation of lithium and the crystal structure of various carbons has been described in detail by Dahn et al. [6]. The charge±discharge curve of graphitized, micron-sized, mesocarbon micro beads (MCMB) is shown in Fig. 5 [7]; the graphite anode sintered at 30008C has a capacity of 400 mAh/g during charging, while on discharge delivers a capacity as low as 290 mAh/g.
This is based on the irreversible reaction (retention) when a thin, lithium ion conductive film (solid electrolyte interface, SEI) is formed on the surface of graphite during the first charge state. In M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 4. Three types of carbon.
Fig. 5. Charge±discharge curve of MCMB heat treated at various temperatures.
graphite, since the charge±discharge voltage is flat, when it is combined with a cathode, the flat voltage discharge pattern is maintained. Such discharge characteristic is a prime requirement in most miniature electrical devices. On the other hand, in hard carbon prepared from polycondensation of oil pitch, lithium is taken in, not only between the carbon layers but also in the crevices between crystal particles (so-called doping place), and as a result, hard carbon accepts more lithium than given by the formula LiC6 and becomes a high capacity anode. The charge±discharge curve of the first cycle of a typical hard carbon (pseudo-isotropic carbon, PIC) is shown in Fig. 6 [8]. Various Fig. 6. Charge±discharge curve of a hard carbon (PIC).
M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 7. Variation of chemical diffusion coefficient of lithium in carbons ((solid square) mesophase pitch-based carbon fiber heated at 30008C; (solid triangle) artificial graphite; (solid circle) natural graphite) in 1 M LiPF6±EC/DEC (1:1) models of the microstructure of hard carbon have been proposed which to possess micro cavities such as pores and voids [9]. Some hard carbons are thought to intercalate or absorb lithium between graphene layers [10,11]. For hard carbons, the irreversibility (retention) of the first cycle as well the potential change that accompanies intercalation is considerable compared with graphite. When hard carbon is used as anode material, the discharge voltage gradually changes with time. Such a discharge pattern permits easy estimation of the remaining capacity of the cell. Also, when several cells with various initial discharge voltages (when unbalance arises in the capacities of cells in a battery) are connected in parallel for the cells with hard carbon as anode, the total battery system maintains a balanced voltage throughout its discharge.
Generally, carbon anodes do not have large energy capacity, and the diffusion of the Li‡ ions between carbon layers is not fast. The chemical diffusion coefficient, D, of Li‡ within carbon is closely related to the power density of battery. Some graphitized carbon fiber has D which is one order of magnitude higher than that in natural graphite (Fig. 7) [12]. Takamura et al. [13] reported recently that covering the entire surface of a graphite fiber with an evaporated film of metallic Ag, Zn or Sn gives a promising technique for improving charge±discharge reaction rates. The most reasonable explanation of the effect is that a newSEI forms on the metal surface. It has been reported that B-doped mesophase pitch-based carbon fiber (MPCF) has an increase of about 10% in its charge capacity with improved cycle efficiency [14].
The search for newer and higher performance anode material leads to the development of alloy anodes based on Sn [15±17] and Si [18,19]. Though the lithium intercalated into Li4M±Li4.4M (M: metal) showed a high capacity of about 800±3000 mAh/g, a stable cycling performance was not obtained as the associated volume change was more than four times during lithium insertion. The use of oxide anode was also proposed and the tin oxides [20,21], InVO4 [22], FeVO4 [22] and Brannerite-type MnV2O6 [23] have been examined. Almost all these oxides change their form from crystalline to amorphous during lithium intercalation. The structure of Brannerite is shown in Fig. 8.
It shows a very high capacity (1000 mAh/g) during first intercalation, but thereafter, shows very high irreversible reaction (retention) and reduced lithiation capacity. The large capacity fade on cycling must be solved for their practice use [23]; stabilization of the structures of these oxide

M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 8. Structure of Brannerite. Mn is represented by the circles and vanadium is located in the center of each octahedron.
materials may be possible by selective partial substitution of the transition metals with other 3.2. Electrolyte materials 3.2.1. Organic electrolyte liquid In lithium ion rechargeable batteries, since the mean charge or discharge voltage remains beyond 3 V ranges, an aqueous electrolyte cannot be used as it would decompose. Hence, inorganic salts (for example, LiClO4) dissolved in organic liquids with a large stability window have been considered as possible electrolytes. To be a good electrolyte, such a non-protonic electrolyte must have good ionic conductivity, high chemical stability, assure safety and be cost-effective. The physical properties of some important organic solvents at room temperature are summarized in Table 2 [24] and their chemical formulae are shown in Fig. 9. Since a lithium ion rechargeable battery usually operates between 20 and ‡608C, solvents with low melting point, high boiling Physical properties of solvents at 258C (unless otherwise noted)Solvent Acetonitrile (AN) g-Butyrolactone (BL) Diethylether (DEE) Dimethylsulfoxide (DMSO) 1,3-Dioxolane (DOL) Ethylene carbonate (EC) Methylformate (MF) Propylene carbonate (PC) Tetrahydrofuran (THE) b Acceptor number.
M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 9. Structural formulae of main organic solvents.
point and lowvapor pressure are highly desirable. The ionic conductivity of Li‡ is proportional to its mobility and the number of mobile ions. Although cyclic carbonic acid esters (carbonate) like PC and EC have high dielectric constants, their viscosity values are high due to the interaction between molecules, which results in deviation of the electric charge of the molecule. Lower viscosity is more desirable for easier movement of the lithium ion. On the other hand, in chain-like esters such as dimethyl carbonate (DMC) and diethyl carbonate (DEC), even though their viscosity and dielectric constant values are quite low, they do not remain much restrictions for the movement of lithium ions between the molecules. Moreover, if the dielectric constant becomes large, the coulombic force between positive and negative sites in a molecule becomes high, imparting a fair degree of ionic dissociation in the neighboring molecule. It is common practice to mix two or more solvents so as to obtain a desired set of properties which is suitable for high performance batteries. The molar conductivity of PC ‡ DMC using LiClO4 or LiPF6 is shown in Fig. 10 [25]. Since the electrolyte based on PC is reported to decompose in contact with graphitized carbon anode [26±28] a suitable solvent mixture (EC/DEC) has been recommended by Yamaki et al. [29]. It has also been proposed that lithium carbonate (Li2CO3) has an important role in providing a protective film on the surface of carbon [26]. The conductivity of mixtures of EC with other solvents is summarized in Table 3 [24].
The systems presented here showvalues around 10 2 S/cm. The conductivity of electrolytes based on mixed carbonate solvents is generally about one half that of electrolytes with DME as the co- solvent, although this depends on the type of lithium salt which is dissolved in it. The effect of electrolyte composition on the charge±discharge characteristics of carbon materials has been examined extensively [30±32]. In usual organic solvents, the water content must be less than 20 ppm, in order to prevent possible chemical side-reaction with lithium. From various investigations, including thermal stability and cyclic voltammetry, LiPF6-based organic electrolyte has become the electrolyte of choice for most commercial handy-type electronic products. The organic electrolyte M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 10. Variation in the molar conductivity of (a) 1 M LiClO4 and (b) LiPF6 with solvent composition in mixed PC ‡ DME at 308C [22].
chosen must be non-inflammable with a wide potential window of more than 5 Von the upper side to satisfy safety requires events and be compatible with a wide range of cathode materials.
3.2.2. Polymer electrolyte When aiming at developing a lithium ion rechargeable battery with full safety features, an all- solid electrolyte becomes very attractive. The development of polymer electrolytes has been in progress for some time, followed by which there has appeared in the market, a polymer-based rechargeable battery (plastic lithium ion battery, PLI) containing a polymer electrolyte membrane.
The membrane acts both as the electrolyte and the separator. The concept is appealing to the battery industry since PLI batteries are expected to be of lowcost and more easily sealed than their liquid electrolyte counterparts. The absence of free liquid implies that PLI batteries can be more easily Conductivity of EC-based electrolytes (EC: co-solvent, 1:1 by volume) at 258CElectrolytic salt (1 mol/dm3) Specific conductivity (mS/cm) M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 11. Mechanism of Li‡ ion conduction in intrinsic (dry) SPE polymer.
compacted in light-weight plastic containers, unlike conventional batteries which require metallic In 1973, Wright [33] made the first measurement of ionic conductivity in polyethylene oxide (PEO) complexed with sodium ions. Armand et al. [34] first examined lithium ion conductivity of mixed lithium salts in PEO and applied it to lithium secondary battery. The conduction mechanism of the lithium ion became clear gradually afterwards. Fig. 11 shows the segmental motion of the polymer chain in the amorphous state [35]. These electrolytes are intrinsic (dry) solid polymer electrolytes (SPE); cation mobility is regulated by lithium ion±polymer interaction involving cation ether oxygen coordination bonds. It has also been explained howlocal relaxation and segmental motion of the host polymer chains is essential to give high conductivity of the electrolyte. Many extensive reviews have been published [36±39]. Some examples for the dry SPE are shown in Fig. 12.
Lower temperatures favor crystallization of the polymer in such dry SPE, and there is no practical use when the associated ionic conductivity falls below 10 8 S/cm at room temperature [34].
Therefore, to develop elastic and self-supporting polymer electrolytes, bridging of PEO chains [40], introduction of side chains [41,42] (comb-type polymer), blending of different kinds of polymers [43] and improvement of lithium salt (supporting salt) have all been tried and polymer electrolytes with conductivity of 10 3±10 4 S/cm at room temperature developed. Recently, Watanabe et al. [44] obtained a polymer electrolyte with high ionic conductivity using high molecular weight (10 5± 10 6) comb-shaped polymers, poly[ethylene oxide-co-2-(2-methoxyethoxy)ethyl glycidyl ether] Fig. 12. Examples of intrinsic (dry) SPE polymers.
M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 13. Arrhenius plots of ionic conductivity (top), and interfacial resistance at the lithium electrode interface (bottom) for P(EO/MEEGE) electrolytes ([Li]/‰OŠ ˆ 0:06).
(P(EO/MEEGE)). The data are shown in Fig. 13 [44]. With increasing MEEGE content, the conductivity increased appreciably due to the fast ion transport assisted by highly flexible ether side chains. The imide salt LiTFSI (LiN(CF3SO2)2) has the highest conductivity in intrinsic dry polymer at present. However, when organic plasticizer is added to the PEO host polymer, it is prone to attack the surface of the aluminum current corrector in the cathode.
Another type of polymer electrolyte known as gel-type solid polymer electrolyte, was first proposed by Feuillard and Perche [45]. This gel-type SPE contains a small amount of suitable organic liquid known as a plasticizer. Plasticizer-containing PAN or PMMA polymer host has been reported to provide high ionic conductivity. The plasticizer effects the polymer in two ways: (1) it makes segmental motion easier and (2) it provides liquid-like character similar to that of a liquid electrolyte. Together, these provide enhanced ionic conductivity in the gel polymer. Plasticizer addition increases choice of host polymers enormously. The relevance of gel polymers to polymer batteries has been fully recognized by Abraham and Alamgir [46,47] and Nagasubramanian et al.
[48]. An example of the host polymer for gel SPE is shown in Fig. 14. The ionically conducting polymer (three-dimensional bridge-type polyalkyleneoxide) presented by SANYO recently is shown in Fig. 15 [49]. For the weight ratio of polymer to electrolyte solution 1:10, the ionic conductivity values obtained were 5  10 3 S/cm at 258C and 2  10 3 S/cm at 108C. Such high ionic conductivity values are almost same as those shown by organic liquid electrolyte solutions. There are reports that the ionic conductivity was slightly improved when inorganic fillers such as g-LiAlO2 and SiO2, etc. were added to the gel-type SPE [50±52]. The ionic conductivity of gel-type SPE is around 10 3 S/cm at room temperature which is far above that of dry SPE and gel-type SPE stands next only to dry solid polymer electrolyte in providing safety to the polymer. Usually, both Li‡ and the related M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 14. Some host polymers of gel SPE.
anion are mobile in amorphous polymer electrolytes [53], and the transference number of Li‡ is less than 0.5. Methods to evaluate the transference number of lithium ions have been proposed by several researchers [54,55]. In general, polymer electrolytes have lower conductivity than liquid electrolytes and cannot deliver high power in room temperature and in particular, low temperature batteries.
However, polymer electrolytes have many excellent points such as ease of battery fabrication in various shapes and better safety than conventional organic liquid electrolytes.
3.3. Cathode materials Whittingham [56] first proposed the feasibility of using an inorganic compound, TiS2 for non- aqueous secondary batteries of high specific and power density during the mid-1970s. This compound has good metallic character and undergoes lithium reversible intercalation. Despite its constant discharge voltage in excess of 2 V with current densities of 1±10 mA/cm2, the difficulties with practical non-aqueous batteries and reversible deposition over lithium negative electrode has restricted commercial LixTiS2/Li cells to button-size units. In 1980, Mizushima et al. [57] proposed using layered LiCoO2 with a-NaFeO2 structure as an intercalation cathode. The measured open circuit voltage of the cell LixCoO2/Li was approximately twice that of LixTiS2 giving a theoretical energy density of 1.1 kWh/kg. It took around 10 years to put LiCoO2 to commercial use. This oxide is structurally more stable than TiS2.
Fig. 15. Temperature dependence of ionic conductivity of gel polymer electrolytes at various weight ratios of polymer to M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 The principal characteristics required for cathode materials are as follows: 1. The discharge reaction should have large negative Gibbs free energy (high discharge voltage).
2. The host structure must have lowmolecular weight and the ability to intercalate large amount of lithium (high energy capacity).
3. The host structure must have high lithium chemical diffusion coefficient (high power density).
4. The structural modifications during intercalation and deintercalation should be as small as possible (long cycle life).
5. The materials should be chemically stable, non-toxic and inexpensive.
6. The handling of materials should be easy.
3.3.1. Materials with a-NaFeO2 structure As already stated, a-NaFeO2-type LiCoO2 is mainly used for the cathode in present lithium ion batteries. It is relatively easy to prepare. The structure is shown in Fig. 16 [58]; LiNiO2, LiCrO2 and LiVO2, also take this structure. These materials contain first trivalent transition metal ions (Ni, Cr, V) having ionic radius smaller than Mn3‡. They are rock salt-structured materials based on a close- packed network of oxygen atoms with Li‡ and M3‡ ions ordering on alternating (1 1 1) planes of the cubic rock salt structure. This (1 1 1) ordering introduces a slight distortion of the lattice to hexagonal symmetry. Hence, LiCoO2 crystallizes in the space group (R3m) with cell constants a ˆ 2:816 AÊ and c ˆ 14:08 AÊ. The lithium ion intercalates into or deintercalates from the van der Waals gap between CoO2 layers (3a site) reversibly. The discharge curve (current density 0.25 mA/ cm2) and the lattice parameter change in Li1 xCoO2 are shown in Fig. 17 [59]. In general, the a-axis changes little, but the c axis changes from 14.1 to 14.6 AÊ with lithium deintercalation. Since a new phase appears near x ˆ 0:5, cycling is performed usually for x-values 0 < x < 0:5.
Fig. 16. Structure of a-NaFeO2.
M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 17. (a) Cell voltage of Li1 xCoO2 (0 < x < 1); (b) hexagonal lattice parameter a; (c) hexagonal lattice parameter c.
As regards mineral resources, only about 10 million t of cobalt deposits are available in the world. However, manganese is about 500 times more abundant (about 5 billion t) than cobalt and is therefore, much cheaper.
LiCoO2 and LiNiO2 form a complete solid solution, LiCoyNi1 yO2 (0 < y < 1) [60]. The charge±discharge behavior of the solid solution was studied first by Delmas and Saadoune [60]. A charge±discharge curve is shown in Fig. 18 [61]. From measurements of 7Li NMR, it is clear that nickel is oxidized during first step, and cobalt in the second step [62,63]. Although the capacity of LiCoO2 is about 130 mAh/g, the capacity of LiCoyNi1 yO2 with part of the cobalt substituted by nickel, increases to about 150 mAh/g but the discharge voltage falls slightly. This solid solution system, especially composition LiNi0.8Co0.2O2, is expected to form a next-generation cathode, after Another layered compound of interest, LiNiO2, has been explored by several workers but suffers from poor thermal stability in its highly oxidized state (Ni3‡/Ni4‡). The Jahn±Teller distortion arising from Ni3‡ (lowspin state) may be a factor responsible for the instability. In fact, the removal of half the lithium from LiNiO2 results in a metastable layered structure Li0.5NiO2 that transforms, on heating to 3008C, to the cubic spinel Li[Ni2]O4. However, constraints in the cubic structure make the spinel structure selective for insertion of Li‡ ions, reducing the Li‡ ion mobility and hence its conductivity. For these reasons, LiNiO2 is of not practical importance at present.
M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 18. Charge±discharge curve: (a) LiCoO2; (b)LiCo1/2Ni1/2O2; (c) LiNiO2 (change density ˆ 0:17 mA/cm2).
M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 19. Zigzag layered structure of LiMnO2. (Large circle) oxygen; (small open circle) manganese; (small solid circle) 3.3.2. Lix(Mn1 xCox) O2 with zigzag layered structure LiMnO2 prepared by usual solid±solid reaction at high temperature slowly transforms into a spinel structure upon cycling. This results in capacity fade, whose extent varies with the cut-off voltage range. This problem has been improved by partial substitution of Mn by other metals.
Recently, Armstrong et al. [64,65] prepared NaMnO2 by solid reaction at high temperature and then synthesized LixMnO2 and Lix(Mn1 yCoy)O2 with zigzag layered structure (Fig. 19) by ion exchange.
When the cut-off voltage and current density were set at 2.5±4.8 V and 0.1 mA/cm2, respectively, for the composition Lix(Mn0.9Co0.1)O2, the capacity exceeded 200 mAh/g (Fig. 20). Although it involves difficult a preparative procedure and shows poor cycle life, it has attracted interest among researchers due to its initial high capacity. A detailed reviewon cathode properties of manganese oxides including lithium manganese oxide spinels, is given by Thackeray [66].
3.3.3. LiMn2O4 spinel system LiMn2O4 and related compounds with spinel structure are prepared from manganese salt, lithium salt and another metal salt when substitution is required, in the stoichiometric ratio, at temperatures around 7508C in air. LiMn2O4 has the cubic spinel structure (Fd3m). Lithium occupies the 8a tetrahedral sites with manganese occupying octahedral (16d) site, and the other octahedral site (16c) is vacant. Oxygen forms a face centered cubic array. The ideal cubic unit cell is shown in Fig. 21(a). When Li‡ diffuses in the structure, first it moves from the 8a site to the neighboring empty octahedral 16c site, then to the next 8a site in such a way that the Li‡ ion takes the diffusion path (8a±16c±8a) (Fig. 21(b)). The angle of the diffusion path is about 1088. There have already been M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 20. Cycle performance of LiMnO2 and LiMn0.9Co0.1O2 [65].
many reports on the charge±discharge characteristics of the cathode LixMn2O4 [67±69]. The open circuit voltage (OCV) curve of LixMn2O4 (0 < x < 2) is shown in Fig. 22 [67]. Charge±discharge cycling in the region (I ‡ II) is possible, while in the flat region (III), the reaction LiMn2O4‡Li ! Li2Mn2O4 proceeds accompanied by a phase transition from cubic spinel to an ordered, tetragonal, NaCl-type structure. Good cycling performance cannot be obtained in the 3 V region (1 < x < 2) because a volume expansion of about 6.4% is associated with the phase transition. The reason for the phase transition is an increase in concentration of Mn3‡ (3d4) ions that form during intercalation of lithium [68], causing a Jahn±Teller distortion. The flat part of the discharge curve in the range x < 0:5 is considered to be a two-phase region with a new cubic phase which exists at x ˆ 0:2. The slight Fig. 21. Structure of spinel (a) unit cell and (b) diffusion path of lithium.
M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 22. Open circuit voltage (OCV) curve of LixMn2O4 (0 < x < 1) at 308C [67].
increase of voltage near x ˆ 0:5 corresponds to an ordered configuration of lithium ions on one half of the tetrahedral 8a sites [70]. The charge±discharge curve in the 4 V region of LiMn2O4 prepared by the dry-type method by Wakihara et al. is shown in Fig. 23 [71]. Although the initial capacity is about 120 mAh/g, by the 100th cycle, it falls to about 60%. They attributed the capacity fade to weak bonding between Mn on 16d octahedral site and the six surrounding oxygen atoms, and calculated some octahedral bonding energies, including Mn±O, on the basis of thermodynamic data [72]. The result is shown in Table 4 [72]. These data were evaluated using Born±Haber cycle assuming a perfect ionic crystal model. The data show that the bonding of Mn±O is the weakest. So, it was thought that stabilization of the 16d octahedral site is essential for cycle life improvement. In order to stabilize the octahedral sub-lattice, attempts have been made by several workers to partially substitute Mn by other transition and non-transition metals.
Fig. 23. Charge±discharge curves of LixMn2O4 in 4 V range at room temperature.
M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Bonding energy of octahedral MO1.5, MO2, and MO1.75 EBE (kJ/mol)MO1.5 MO1.75 (0.5M2O3.5) The first charge±discharge characteristics of LixM1/6Mn11/6O4 (M ˆ Cr, Co, Al and Ni) at room temperature are shown in Fig. 24 [71]. In this formula, M1/6 implies that the metal M substitutes 1/12 of manganese at 16d sites. The capacity fade is minimized in all the substituted spinels. The initial discharge capacities (>110 mAh/g) of substituted spinels LixM1/6Mn11/6O4 (M ˆ Cr, Co, Al) are higher than the initial capacity (95 mAh/g) of LixNi1/6Mn11/6O4. The nickel-substituted spinel differs from other spinels LixM1/6Mn11/6O4 (M ˆ Cr, Co, Al) due to its lower valence (Ni2‡). The substituents Cr3‡ and Co3‡ directly substitute for Mn3‡, whereas for Ni2‡ substitution at the Mn3‡ position, charge compensation increases the amount of Mn4‡. Reduction in the amount of Mn3‡ causes reduced capacity in nickel-substituted spinel. However, there is little change in the discharge voltage with different substituent elements. The variation of energy density with cycle for LiM1/6Mn11/6O4 (M ˆ Cr, Co, Ni, Al) is shown in Fig. 25. The cells with substituted LiMn2O4 showbetter cycle characteristics than the one with undoped LiMn2O4. Actually, the electrochemical cycling properties on partial substitution of manganese by other metals has been reported by many researchers [68,71,73,74]. It has been reported that only about 10 mol% of substitution for Mn3‡ in the 16d octahedral site is sufficient to effect improved cycle performance. It has been observed that Mn3‡ (3d4) is a Jahn±Teller ion and causes a phase transition from cubic to tetragonal (1.4 kJ/mol) in LiMn2O4 near 08C [75], but the enthalpy of the phase transition in partially-substituted spinels gradually decreases with increasing amount of substituent, and the phase transition is completely suppressed with 10±20 mol% of substitution in octahedral Mn site [74]. A DSC curve of LiMyMn2 yO4 obtained by our group is presented in Fig. 26 [74]. For high cycle performance, it is desirable that a phase transition does not occur in the middle of an intercalation reaction.
Fig. 24. The first charge±discharge curve of LiM1/6Mn11/6O4 at room temperature.
M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 25. The relationship between energy capacity and cycle number for LixMn2O4 at room temperature.
As mentioned above, LiMn2O4 and LiMyMn2 yO4 can be prepared easily by carrying out high temperature calcination in air. These oxides are cost-effective and environmentally benign cathodes for applications from ranging miniature batteries to large-scale batteries for EV and energy storage applications. However, the capacity fade on cycling is much higher if cycling is performed at higher temperature, 50±708C. Dissolution of manganese into the organic electrolyte liquid during cycling appears to be one of the causes [76,77]. In addition, the deposition of metallic Mn on the anode surface has also been mentioned as a possible reason for the large capacity fade. In our recent research [78,79], we found that dissolution of Mn is very low in doped LiMn2O4, even at 508C (Fig. 27) [78,79]. Among these dopants, Co seems to be effective and a good cycling performance has been obtained (Fig. 28) [78,79]. When Mn dissolves into the liquid electrolyte, it exists as Mn2‡ as shown by ESR [80] and in situ XANES [81] measurements. It is also presumed that some Li‡ from the electrolyte may enter the 16d octahedral sites, causing some Mn3‡ in the octahedral site to oxidize to Mn4‡. The problem of dissolution of Mn at high temperature must be solved for practical applications because the cell temperature in EV could rise as high as 708C during summer. To overcome Mn dissolution, besides partial substitution, methods such as introduction of defects into Fig. 26. DSC curves of LiM1/6Mn11/6O4 (M ˆ Al1/6, Al1/12Co1/12, Al1/12Cr1/12) and LiMn2O4 at temperatures between 235 and 320 K on heating.
M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 27. Dissolved Mn concentration in the electrolyte for LiM1/6Mn11/6O4 (M ˆ Al, Ni, Co) at 508C.
the structure, use of a polymer electrolyte membrane and reexamination of the quality of the electrolyte liquid have all been considered.
In situ XANES and EXAFS spectra sometimes give effective information on the change in valence state of metals during electrochemical cycling. Fig. 29 shows Mn K-edge in situ XANES spectra of LixMn2O4 during deintercalation of lithium [82]. The absorption spectrum shifts to higher energy with deintercalation of lithium, indicating that Mn3‡ in the spinel oxidizes to Mn4‡.
Considering the possible use of spinel-based cathodes, one may conclude that materials from the LiMyMn2 yO4 (M ˆ Co, Cr, Al, Ni, Li) family will be the most likely candidates for cathodes in the next generation of power sources.
3.3.4. Other oxides Crystalline V2O5 with a shear structure [83,84], amorphous a-V2O5 [85], a-V2O5±B2O3, and a- V2MoO8 [86] have all been considered. The average discharge voltage is around 2.5 V and is not as high as that of LiCoO2 and LiMyMn2 yO4. Since there is also capacity fade with cycling, these materials have not been exploited for large-scale development. Recently, it has been reported that LixFeyOz with (Li‡/Fe3‡ ˆ 0:69) prepared at 2008C had a capacity of about 140 mAh/g at an average discharge voltage of 2 V during cycling [87]. Since the cost of Fe is very cheap, it may become an attractive candidate in the future. An olivine (M2SiO4)-based cathode using the crevice Fig. 28. Cycle performance for Li LiM1/6O4 (M ˆ Al, Ni, Co) and Li LiMn2O4 cells at 508C.
M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 29. Mn K-edge in situ XANES spectra of LixMn2O4 (0 < x < 1).
between MO4 tetrahedra has been reported recently by Padhi et al. [88]. They considered four iron phosphates, Li3Fe2(PO4)3, LiFeP2O7, Fe4(P2O7)3 and LiFePO4 and found that LixFePO4 delivered 130 mAh/g at an average cell voltage of 3.3 V. In order to increase the discharge voltage, LiCoPO4 instead of LiFePO4 has been prepared, and its cycling behaviors examined [89]. Although the average discharge voltage increased to 4.5 V, satisfactory cycleability could not be observed because of gradual decomposition of the organic electrolyte. There are no stable organic electrolyte liquids that work well at discharge±charge voltages above 4.5 V at present, and further work to obtain improved electrical characteristics of organic electrolytes is needed.
4. Manufacture process So far, the diverse research activities that lead to the development of cathode and anode active materials, separator, electrolyte, current collector (metal foil; cathode is aluminum foil and anode Cu foil), material optimization and possible materials for use in a lithium ion battery have been The manufacturing processes consist of (1) mixing the cathode or anode materials with binder and conductive additive, (2) painting on the current collector (metal foil), (3) drying and (4) pressing (Fig. 30). The next step involves (1) assembly of cathode, anode and the simultaneous rolling-up with separator, (2) electrode insertion, (3) electrolyte injection into the battery case and (4) sealing.
In the electrode manufacturing process, a mixture of active materials with conductive additive such as acetylene black or ketjen black, and a binder such as Teflon or polyvinylidenefluoride (PVdF) dissolved in n-methyl-2-pyrrolidinone (NMP) are made in the form of a paste. The paste is painted over both sides of metal foil, dried and roll-pressed. Then it is cut to the desired width. In the case of polymer lithium batteries [38], after carrying out direct polymerization of the gel electrolyte sheet, roll-press is carried out over the cathode and anode sheets which are then cut a suitable size,

M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 Fig. 30. Schematic processing of cathode film for lithium ion battery.
Fig. 31. Schematic manufacture processing of lithium polymer battery.
wrapped in aluminum lamination film and heat welded at the edges (Fig. 31). Finally, the battery is checked for short-circuit over 2±4 weeks. A protective thin film (SEI) will form at the anode±carbon interface during this period. Shipment inspection has to be conducted and the products shipped.
Many battery manufacturing companies worldwide have announced their involvement in large-scale production of lithium secondary batteries. However, the level of information released is limited and thus, it is impossible to evaluate the status of the different batteries at this stage.
A lithium ion rechargeable battery appears to occupy a prime position in the market place and it took almost a decade to reach this strong position. Attempts have been made to eliminate the use of detrimental heavy metal, to avoid environmental hazards, and to make material improvements to give cost-effectiveness. An immense amount of research has been carried out on the cathode, anode and electrolyte and simplification of the manufacturing process. Newmaterials developments and newtechnical developments are expected to help realize the goal of large-size batteries to power electric vehicle as well producing a range of small batteries for the next millennium.
M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 The author wishes to thank Dr. A. Veluchamy for his valuable discussion and Dr. Y. Uchimoto, Dr. H. Ikuta and Dr. S. Dong for their experimental work and helpful discussion. Financial support from following source: Grant-in-Aid for Scientific Research on Priority Areas (B) (No. 11229201) Fundamental Studies for Fabrication of All Solid State Ionic Devices'' from Ministry of Education, Science, Sports and Culture in Japan.
[1] Y. Nishi, Performance of the first lithium ion battery and its process technology, in: M. Wakihara, O. Yamamoto (Eds.), Lithium Ion Batteries, Kodansha, Tokyo; Wiley/VCH, Weinheim, 1998, p. 181.
[2] K. Nakajima, Y. Nishi, Advanced Li-ion batteries, in: T. Osaka, M. Datta (Eds.), Energy Stage Systems for Electronics, Gordon and Breach, Amsterdam, 2000, p. 109.
[3] Z.X. Shu, R.S. McMillan, J.J. Murry, J. Electrochem. Soc. 140 (1993) 922.
[4] J.M. Tarascon, D. Guyomard, Electrochim. Acta 38 (1993) 1221.
[5] T. Ohzuku, Y. Iwakoshi, K. Swai, J. Electrochem. Soc. 140 (1993) 2490.
[6] J.R. Dahn, A.K. Sleigh, H. Shi, J.N. Reimers, Q. Zhong, B.M. Way, Electrochim. Acta 38 (1993) 1179.
[7] N. Takami, in: T. Takamura (Ed.), Battery Handbook, Asakura Shoten, Tokyo, 1996, p. 787.
[8] N. Sonobe, M. Ishikawa, T. Iwasaki, in: Proceedings of the Abstracts of 35th Battery Symposium, Nagoya, Japan, 1994, p. 47.
[9] A. Mabuchi, K. Tokumitsu, H. Fujimoto, T. Kasuh, J. Electrochem. Soc. 142 (1995) 1041.
[10] K. Sato, M. Noguchi, A. Demachi, N. Oki, M. Endo, Science 264 (1994) 556.
[11] T. Zheng, J.S. Xue, J.R. Dahn, Chem. Mater. 8 (1996) 389.
[12] N. Takami, A. Satoh, M. Hara, T. Ohasaki, J. Electrochem. Soc. 142 (1995) 1090.
[13] T. Takamura, K. Sumiya, J. Suzuki, C. Yamada, K. Sekine, J. Power Sources 81/82 (1999) 368.
[14] M. Endo, C. Kim, T. Karaki, Y. Nishimura, M.J. Matthews, S.D.M. Brown, M.S. Dresselhaus, Carbon 37 (1999) 561.
[15] J.O. Besenhard, J. Yang, M. Winter, J. Power Sources 68 (1997) 87.
[16] J. Yang, Y. Takeda, N. Imanishi, T. Ichikawa, O. Yamamoto, Extended Abstracts, SSI-12, A-31, Halkidiki, Greece, 1999, p. 88.
[17] M. Nagayama, T. Morita, H. Ikuta, M. Wakihara, M. Takano, S. Kawasaki, Solid State Ionics 106 (1998) 33.
[18] M. Wakihara, T. Morita, A. Modeki, H. Ikuta, Extended Abstracts, SSI-12, A-34, Halkidiki, Greece, 1999, p. 92.
[19] H. Kim, Y.-J. Kim, D.-G. Kim, H.-J. Sohn, T. Kang, B. Park, Extended Abstracts, IMLB10, Como, Italy, 2000, p. 78.
[20] Y. Idota, T. Kubota, A. Matsuji, Y. Maekawa, T. Miyasaka, Science 276 (1997) 1397.
[21] P. A. Connor, F. Belliard, M. Behn, J.T.S. Irvan, Extended Abstracts, SSI-12, A-29, Halkidiki, Greece, 1999, p. 85.
[22] S. Denis, E. Baudrin, F. Orsini, G. Ouvrard, M. Toboul, J.-M. Tarascon, J. Power Sources 81/82 (1999) 79.
[23] S.-S. Kim, H. Ikuta, M. Wakihara, Solid State Ionics 139 (2001) 57.
[24] M. Morita, M. Ishikawa, Y. Matsuda, Organic electrolytes for rechargeable lithium ion batteries, in: M. Wakihara, O. Yamamoto (Eds.), Lithium Ion Batteries, Kodansha, Tokyo; Wiley/VCH, Weinheim, 1998, p. 156.
[25] M. Matsuda, M. Morita, F. Tachihara, Bull. Chem. Soc. Jpn. 59 (1986) 1967.
[26] Y. Ein-Eei, B. Markovsky, D. Aurback, Y. Carmeli, H. Yamin, S. Luski, Electrochim. Acta 39 (1994) 2559.
[27] D.E. Irish, Z. Deng, M. Odziemkowski, J. Power Sources 54 (1995) 28.
[28] P. Liu, H. Wu, J. Power Sources 56 (1995) 81.
[29] M. Arakawa, J. Yamaki, J. Electronal. Chem. 219 (1987) 273.
[30] D. Aubach, Y. Ein-Eei, B. Markovski, A. Zaban, S. Luski, Y. Carmeli, H. Yamine, J. Electrochem. Soc. 142 (1995) [31] T. Iijima, K. Suzuki, Y. Matsuda, Synth. Met. 73 (1995) 9.
[32] A. Ohta, H. Koshina, H. Okuno, H. Murai, J. Power Sources 54 (1995) 6.
[33] P.V. Wright, Br. Polym. J. 7 (1975) 319.
[34] M.B. Armand, J.M. Chabano, M.J. Duclot, in: P. Vashista, J.N. Mundy, G.K. Shenoy (Eds.), Fast Ion Transport in Solids, North-Holland, NewYork, 1979, p. 131.
[35] D.F. Shriver, G.C. Farrington, Chem. Eng. News 20 (1985) 42.
[36] B. Scrosati, R. Neat, in: B. Scrosati (Ed.), Publication of Electroactive Polymers, Chapman & Hall, London, 1993, [37] M. Armand, J.Y. Sanchez, M. Gauthier, Y. Coquette, in: J. Lopkowski, P.N. Ross (Eds.), Electrochemistry of Novel Materials, VCH, Weinheim, 1994, p. 65.
M. Wakihara / Materials Science and Engineering R33 (2001) 109±134 [38] B. Scrosati, Lithium ion plastic batteries, in: M. Wakihara, O. Yamamoto (Eds.), Lithium Ion Batteries, Kodansha, Tokyo; Wiley/VCH, Weinheim, 1998, p. 218.
[39] F.B. Dias, L. Plomp, J.B. Veldhuis, J. Power Sources 88 (2000) 169.
[40] J.F. Le Nest, S. Callens, A. Gandini, M.B. Armand, Electrochim. Acta 37 (1992) 1585.
[41] P.M. Blonsky, D.F. Schriver, P. Austin, H.R. Allcock, Solid State Ionics 18/19 (1986) 258.
[42] A. Nishimoto, M. Watanabe, Y. Ikeda, S. Kohjiya, Electrochim. Acta 43 (1998) 1177.
[43] S. Yokoyama, T. Kobayashi, Y. Kato, H. Ikuta, M. Wakihara, Extended Abstracts, IMLB10, Como, Italy, 2000, [44] M. Watanabe, T. Endo, A. Nishimoto, K. Miura, M. Yanagida, J. Power Sources 81/82 (1999) 786.
[45] G. Feuillard, P. Perche, J. Appl. Electrochem. 5 (1975) 63.
[46] K.M. Abraham, M. Alamgir, J. Electrochem. Soc. 136 (1990) 1657.
[47] K.M. Abraham, M. Alamgir, J. Power Sources 44 (1993) 195.
[48] G. Nagasubramanian, A.I. Attia, G. Halpert, J. Appl. Electrochem. 24 (1994) 298.
[49] S. Narukawa, I. Nakane, Extended Abstracts, IMLB10, Como, Italy, 2000, p. 38.
[50] J.E. Weston, B.C. Steele, Solid State Ionics 7 (1982) 75.
[51] F. Croce, B. Scrosati, J. Power Sources 43 (1993) 9.
[52] M.C. Borghini, M. Mastragostino, S. Passerni, B. Scrosati, J. Electrochem. Soc. 142 (1995) 2118.
[53] P.G. Bruce, C.A. Vincent, Solid State Ionics 40/41 (1990) 607.
[54] P.R. Sorensen, T. Jacobsen, Electrochim. Acta 27 (1982) 1671.
[55] P.M. Blonsky, D. Shriver, P. Austin, H.R. Allcock, 18/19 (1986) 258.
[56] M.S. Whittingham, Prog. Solid State Chem. 12 (1978) 41.
[57] K. Mizushima, P.C. Jones, P.J. Wiseman, J.B. Goodenough, Mater. Res. Bull. 15 (1980) 783.
[58] T.A. Hewston, B.L. Chamberland, J. Phys. Chem. Solids 48 (1987) 97.
[59] K. Hoshino, T. Murakami, A. Atsuka, Y. Ozaki, S. Watanabe, Y. Takahashi, Nat. Tech. Rep. 40 (1994) 31.
[60] C. Saadoune, I. Delmas, Solid State Ionics 53±56 (1992) 370.
[61] T. Ohzuku, A. Ueda, M. Nagayama, Y. Iwakoshi, H. Komori, Electrochim. Acta 38 (1993) 1159.
[62] M. Menetrier, A. Rougier, C. Delmas, Solid State Commun. 90 (1994) 439.
[63] C. Marichal, J. Hirschinger, P. Granger, M. Menetrier, A. Rougier, C. Delmas, Inorg. Chem. 34 (1995) 1773.
[64] A.R. Armstrong, P.G. Bruce, Nature 381 (1996) 499.
[65] A.R. Armstrong, H. Huang, R.A. Jennings, P.G. Bruce, J. Mater. Chem. 8 (1998) 255.
[66] M.M. Thackeray, Prog. Solid State Chem. 25 (1997) 1.
[67] T. Ohzuku, M. Kitagawa, T. Hirai, J. Electrochem. Soc. 137 (1990) 769.
[68] J.M. Tarascon, E. Wang, F.K. Shokoohi, W.R. Mckinnon, S. Colson, J. Electrochem. Soc. 138 (1991) 2859.
[69] J.M. Tarascon, D. Guyomard, J. Electrochem. Soc. 138 (1991) 2864.
[70] J.B. Goodenough, M.M. Thackeray, W.I.F. David, P.G. Bruce, Rev. de Chim. Miner. 21 (1984) 435.
[71] G. Li, H. Ikuta, T. Uchida, M. Wakihara, J. Electrochem. Soc. 143 (1996) 178.
[72] M. Wakihara, G. Li, H. Ikuta, Cathode active materials with a three-dimensional spinel framework, in: M. Wakihara, O. Yamamoto (Eds.), Lithium Ion Batteries, Kodansha, Tokyo; Wiley/VCH, Weinheim, 1998, p. 26.
[73] G. Pistoia, G. Wang, Solid State Ionics 66 (1993) 135.
[74] D. Song, H. Ikuta, T. Uchida, M. Wakihara, Solid State Ionics 117 (1999) 151.
[75] A. Yamada, M. Tanaka, Mater. Res. Bull. 30 (1995) 715.
[76] Y. Xia, M. Yoshio, J. Electrochem. Soc. 143 (1996) 825.
[77] D.H. Jang, Y.J. Shin, S.M. Oh, J. Electrochem. Soc. 143 (1996) 825.
[78] D. Song, H. Ikuta, M. Wakihara, Electrochemistry 68 (2000) 460.
[79] M. Wakihara, Extended Abstracts, IMLB10, Como, Italy, 2000, p. 16.
[80] S. Kanamura, H. Naito, T. Yao, Z. Takehara, J. Mater. Chem. 6 (1996) 33.
[81] I. Nakai, Private Communication.
[82] I. Nakai, Y. Shiraishi, F. Nishikawa, Spectrochim. Acta B54 (1999) 143.
[83] J.M. Cocciantelli, J.P. Doumerc, M. Pouchard, M. Broussely, J. Labat, J. Power Sources 34 (1991) 103.
[84] J.M. Cocciantelli, M. Menetrier, C. Delmas, J.P. Doumerc, M. Pouchard, P. Hagenmuller, Solid State Ionics 50 [85] C. Delmas, S. Brethes, M. Menetrier, J. Power Sources 34 (1991) 113.
[86] M. Wakihara, Manganese spinel oxides and vanadium oxides for cathode active materials, in: Z. Ogumi (Ed.), The Latest Technology of the NewSecondary Battery, CMC, Tokyo, 1997, p. 31.
[87] J. Kim, A. Manthiram, J. Electrochem. Soc. 146 (1999) 4371.
[88] A.K. Padhi, K.S. Nanjundaswamy, C. Masquelier, S. Okada, J.B. Goodenough, J. Electrochem. Soc. 144 (1997) [89] S. Okada, H. Arai, J. Yamaki, Denki Kagaku 65 (1997) 802.

Source: http://www.hulliger.dcb.unibe.ch/pages/teaching/anorg_chemie/scripts/lithium_batteries.pdf