Optimised positive electrode material for lithium cell batteries, method for the production thereof, electrode, and battery for implementing said method

The present invention relates to a novel compound with a spinel structure, of the mixed oxide type based on Ni, Mn and Li, non-stoichiometric and having a clearly defined mesh parameter.

Such a compound has optimised properties, in terms of stability and electrochemical performance.

In consequence, it is advantageously used in electrodes, batteries and cell batteries containing lithium.

Lithium cell batteries are increasingly used as self-contained energy sources, particularly for portable equipment, where they progressively tend to replace nickel-cadmium (NiCd) and nickel-metal hydride (NiMH) batteries. This development stems from the fact that the performance of lithium cell batteries, in terms of power density (Wh/kg, Wh/l), is substantially superior to that of the two abovementioned technologies.

The active electrode compounds used in these batteries mainly have the formulas LiCoO₂, LiNiO₂ and LiMn₂O₄ for the positive electrode, and carbon (graphite, coke, etc.) for the negative electrode. The theoretical and practical capacitances of these compounds are respectively 275 mAh/g and 140 mAh/g for LiCoO₂ and LiNiO₂, and 148 mAh/g and 120 mAh/g for LiMn₂O₄, for an operating voltage close to 4 volts with respect to lithium metal.

The existing compound LiNi^II_0.5Mn^IV_1.5O₄, having a spinel structure, is electrochemically active reversibly at a potential of 4.7-4.8 V/Li/Li⁺. Its theoretical specific capacitance is 147 mAh/g. This gives it a theoretical power density of about 700 Wh/Kg/Li⁺/Li.

Despite numerous and varied studies on materials derived from LiMn₂O₄, few data are available on the compound LiNi_0.5Mn_1.5O₄. The main data available concern the optimisation of its synthesis and the electrochemical behaviour and performance of this material, as positive electrode for a lithium battery.

It should be noted that the term “electrochemical performance” of a material pertains to two different concepts:

the capacity to respond to high current (power) densities; and

the capacity to have a long cycling service life (high and stable capacitance in cycling).

These two properties can be obtained simultaneously for a given compound, but this simultaneous feature is not systematically obtained. Depending on the target application, one and/or the other of these properties is therefore desirable.

Document JP-A-8 217452 discloses a material with a spinel structure, with the formula LiNi_xMn_2-xO₄, in which x is between 0 and 0.5 inclusive. It is reported that this compound has good cycling properties (without specification), due to the use, as precursor, of a manganese oxide MnO₂ having a specific surface area of 150 to 500 m²/g.

Document FR 2 831 993 relates to the doping of a compound with the general formula LiMn_2(x+y)M_xM′_yO₄ by element M′, with M═Ni or Co and M′═Ti, Al, Co, Mo, x and y strictly positive. This results in a material which operates well at high potential, with high capacitance and good cycling properties.

Furthermore, other studies, such as the publication of Sun et al. (Electrochimica Acta 48(2003) 503), recommend a surface treatment of the material (based on ZnO for example) to improve the cycling performance.

Even more recently, Kim et al. (Chem. Mat. 16(2004), 906) showed that a derivative of LiNi_0.5Mn_1.5O₄, non-stoichiometric, with a spinel structure Fd-3m, and a formula Li_1.00Ni_0.48Mn_1.5O_3.92 and mesh parameter 8.172 Å, compared with its purely stoichiometric homologue with a mesh parameter of 8.166 Å, has a better insertion kinetics, and hence better response at high current density.

Thus, values of 137 mAh/g have been obtained for the two compounds at C/7, while at 3C, a value of 110 mAh/g was measured for the non-stoichiometric compound, against 60 mAh/g for the stoichiometric compound. These better properties have been attributed by the authors to the difference in stoichiometry alone, although other crucial factors, particularly the morphology and grain size, have not been taken into account.

Patent application JP 2002 158007 A also relates to the compound with the formula LiNi_0.5Mn_1.5O₄, with mesh parameters of 8.18 Å or more. It is certainly demonstrated that the mesh parameters of between 8.174 and 8.179 Å can be obtained, but only after heat treatment, and hence a modification of the composition. Furthermore, the Mn^III/Mn^IV distribution in this compound remains undetermined.

Methods for synthesising this class of compounds have also been described in (Perentzis et al., J. SOLID STATE ELECTROCHEM. 8(1) 2003, 51-54; Ito et al., JOURNAL OF POWER SOURCES 119-121 (2003), 733-37).

The present development of lithium cell batteries reveals the need to identify novel compounds, serving as a positive electrode material and having optimised electrochemical properties, particularly in terms of cycled capacitance and the cycling stability of this capacitance.

In the context of the invention, the Applicant has obtained compounds with a spinel structure, of the mixed oxide type based on nickel, manganese and lithium, non-stoichiometric, with novel and clearly defined formulas and morphology, having the desired properties.