Doped spinel, method for the production thereof, the use thereof and lithium-ion battery

Abstract

A doped spinel comprising the formula:
Li.sub.1±wMe1.sub.vMe2.sub.x-vMn.sub.2-x-yTiyO.sub.4-zF.sub.z
where, 0≦w<1, 0.3<x≦0.7, 0.3≦v<0.7, x>v, 0.0001≦y≦0.35, and 0.0001≦z≦0.3. Me1 is a metal selected from a group of elements consisting of Cr, Fe, Co, Ni, Cu, and Zn. Me2 is a metal selected from a group of elements consisting of Ni, Fe, Co, Mg, Cr, V, Ru, Mg, Al, Zn, Cu, Cd, Ag, Y, Sc, Ga, In, As, Sb, Pt, Au, and B.

Claims

1. A doped spinel comprising the formula
Li.sub.1±wMe1.sub.vMe2.sub.x-vMn.sub.2-x-yTiyO.sub.4-zF.sub.z where, 0≦w<1, 0.3<x≦0.7, 0.3≦v<0.7, x>v, 0.0001≦y≦0.35, 0.0001≦z≦0.3, Me1 is a metal selected from a group of elements consisting of Cr, Fe, Co, Ni, Cu, and Zn, and Me2 is a metal selected from a group of elements consisting of Ni, Fe, Co, Mg, Cr, V, Ru, Mg, Al, Zn, Cu, Cd, Ag, Y, Sc, Ga, In, As, Sb, Pt, Au, and B, wherein, a specific capacity of the doped spinel from 1 to 50 cycles did not decrease in a voltage range of from 2.0 to 5.0 V, and the doped spinel is prepared by a process consisting of: providing a solution of a metal-organic and/or inorganic Li, Mn, Me, Me1 and/or Me2 compound(s) as a precursor; spray drying the precursor with trifluoroacetic acid so as to obtain a sprayed precursor; calcining the sprayed precursor so as to provide the doped spinel in a form of a pulverulent material; milling the pulverulent material in ethanol so as to provide a suspension; adding a lithium compound and a titanium alkoxide in an ethanolic solution to the suspension so as to provide a resulting material; spray granulating the resulting material so as to obtain a spray-granulated material; and thermally converting the spray-granulated material so as to obtain a Ti-doped oxyfluoride spinel.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The present invention is described in greater detail below on the basis of embodiments and of the Figures in which:

(2) FIG. 1 shows specific discharging capacities of the samples A to E at 23° C. and different voltage ranges;

(3) FIG. 2 shows charging and discharging curves of the samples A to E at 23° C. in the voltage range from 5.0 V to 2.0 V after 10 cycles;

(4) FIG. 3 shows charging and discharging curves of the samples A to E at 23° C. in the voltage range from 5.0 V to 2.0 V after 100 cycles;

(5) FIG. 4 shows the coulombic efficiency of the samples A to E at 23° C. and different voltage ranges;

(6) FIG. 5 shows specific energy densities of the samples A to E at 23° C. and different voltage ranges;

(7) FIG. 6 shows efficiency of the samples A to E at 23° C. and different voltage ranges; and

(8) FIG. 7 shows specific discharging capacities of the samples A to E at 45° C. and different voltage ranges.

DETAILED DESCRIPTION

(9) The material according to the present invention is a Ti-doped oxyfluoride-lithium-manganese spinel. In an embodiment, the spinel can, for example, have the composition:
Li.sub.1±wMe.sub.xMn.sub.2-x-yTi.sub.yO.sub.4-zF.sub.z,

(10) where, in the composition indicated,

(11) 0≦w≦1,

(12) 0≦x≦0.7,

(13) y≦0.3,

(14) z≦0.3, and

(15) Me is at least one element selected from the group of elements consisting of {Li, Ni, Fe, Co, Mg, Cr, V, Ru, Mg, Al, Zn, Cu, Cd, Ag, Y, Sc, Ga, In, As, Sb, Pt, Au, B}. The value range selected for w corresponds to the maximum possible stability range of a normal, i.e., non-inverse, Li-transition metal spinel. At values of z>0.3, the specific capacity decreases to an increasing extent for industrially relevant power densities.

(16) In an embodiment of the present invention, a doped high-voltage spinel which has the formula:
Li.sub.1±wMe1.sub.vMe.sub.2-vMn.sub.2-x-yTi.sub.yO.sub.4-zF.sub.z,

(17) where 0.3≦v≦0.7, in which Me.sub.x is replaced by Me1.sub.vMe2.sub.x-v, Me1 is selected from the group of elements consisting of {Cr, Fe, Co, Ni, Cu, Zn} and Me2 is selected from the group of elements consisting of {Li, Ni, Fe, Co, Mg, Cr, V, Ru, Mg, Al, Zn, Cu, Cd, Ag, Y, Sc, Ga, In, As, Sb, Pt, Au, B}, can, for example, be provided.

(18) In an embodiment of the present invention, the element Ni can, for example, be selected for Me1, with Me2, for example, being selected from the group consisting of {Cr, Fe, Co, Ni, Cu, Zn, Mg, V, Ru, Al}, for example, Cr, Fe or Ru, for example, Fe. In an embodiment, Me2=Fe, 0≦(x-v)≦0.4, 0.0005≦y≦0.2 and 0.0005≦z≦0.3. For values of (x-v)>0.35, the specific capacity in the voltage range>3.5 V of LiNi.sub.0.5-(x-v)Mn.sub.1.5Fe.sub.(x-v)O.sub.4 decreases again as described in Alcantara et al., Journal of The Electrochemical Society 152, pp. A13-A18 (2005). For values of y>0.35, the specific capacity in the voltage range>3.5 V of LiNi.sub.0.5Mn.sub.1.5-yTi.sub.yO.sub.4 decreases again as described in Alcantara et al., Chemistry of Materials 15, pp. 2376-2382 (2003).

(19) A spinel doped according to the present invention is obtained in good quality by a multistage preparative process. Firstly, a solution of Li, Mn, Me, Me1 and/or Me2 metal compounds and trifluoroacetic acid is spray dried. Subsequent calcination in a first furnace operation forms the spinel, for example, still without Ti doping. To introduce the Ti doping, the pulverulent material can, for example, be milled and spray-granulated as an ethanolic suspension in which a lithium salt and a Ti alkoxide has been dissolved, for example, as a sol. The product obtained in this way is converted into a Ti-doped oxyfluoride spinel according to the present invention in a second furnace operation.

(20) The spinel doped according to the present invention makes it possible to cycle a spinel of the class LiMn.sub.2O.sub.4 or a high-voltage spinel of the class LiMe1.sub.vMn.sub.1.5-vO.sub.4 where Me1={Cr, Fe, Co, Ni, Cu, Zn}, 0.3≦v≦0.7, having significantly more than one lithium per formula unit over many cycles in a stable manner. For high cycling stability, specific capacities above 150 mAh/g were measured, which corresponds to the de/intercalation of one lithium per formula unit. Although improved resistance to the electrolyte and good cycling properties are known for doping with iron, fluorine-doped LiNi.sub.0.5Mn.sub.1.5O.sub.4 has improved resistance to hydrogen fluoride and improved heat resistance and, finally, fluorine- and iron-doped LiNi.sub.0.5Mn.sub.1.5O.sub.4 improves the reversibility of deep discharge down to 2.0 V, the class of materials according to the present invention of Ti-doped oxyfluoride spinels or high-voltage spinels combines, firstly, the above-mentioned positive influences and, in addition, achieves a higher energy density as a result of the de/intercalation of more than one lithium per formula unit. This also applies at elevated temperature in combination with a liquid electrolyte, which indicates improved aging stability.

(21) In a range which is given by a doped spinel of the formula Li.sub.1±wNi.sub.vFe.sub.xMn.sub.2-x-yTi.sub.yO.sub.4-xF.sub.z where:

(22) 0≦w≦1,

(23) 0.3≦v≦0.7,

(24) 0≦x≦0.4,

(25) 0.0005≦y≦0.3, and

(26) 0.0005≦z≦0.3,

(27) it is found that the doping element Fe increases electronic conductivity. Iron-doped LiNi.sub.0.5Mn.sub.1.5O.sub.4 therefore gives better performance than undoped material. The specific capacity of iron-doped LiNi.sub.0.5Mn.sub.1.5O.sub.4 is higher than that of undoped material at a medium discharge rate, i.e., over a period of 2 hours, not only in the high-voltage range but also in the 3V range. Fluorine doping improves the stability to hydrogen fluoride formation in the electrolyte and increases the thermal stability of the material since the spinel lattice is stabilized by the introduction of fluoride ions. The combination of iron doping and fluorine doping thus results overall in increased thermal stability and a greater cycling stability in the case of deep discharge.

(28) In the case of deep discharge, the transformation of the cubic spinel lattice into a tetragonal lattice at medium discharging rates is hindered. As a result, the Fe—F-doped material is more stable to cycling but, due to the lower de/intercalation of Li in the 3V range, the specific capacity and energy density during cycling at 2.0-5.0 V increase only little compared to the capacity achieved during cycling at 3.5-5.0 V. Only doping with a third element allows the positive properties of the single doping, i.e., higher specific capacity as a result of iron doping, improved structural stability and improved resistance to the electrolyte as a result of fluorine doping, to be maintained in combination. The titanium-, fluorine- and iron-doped spinel thus displays an increased energy density combined with comparatively high cycling stability and thermal stability and is charged or discharged over the entire voltage range of 2.0-5.0 V in combination with a lithium anode.

(29) The spinel or high-voltage spinel doped according to the present invention is suitable first and foremost for use as cathode material, for example, as an intercalation material, in lithium ion cells, for example, in lithium metal cells having lithium as an anode.

(30) In an embodiment, the present invention provides a lithium ion battery which has at least one lithium ion cell whose cathode material contains a spinel or high-voltage spinel doped according to the present invention.

(31) The present invention will be illustrated below with the aid of the working examples and the Figures.

(32) As a working example, titanium-, fluorine- and iron-doped spinels were prepared and used as the cathode material. Doping was carried out via a modified sol-gel process, starting from aqueous metal salt solutions with addition of trifluoroacetic acid, by which means the following samples A to E were prepared.

EXAMPLES

Comparative Samples A to C

(33) Sample A: Li.sub.1.00Ni.sub.0.52Ti.sub.0.02Mn.sub.1.46O.sub.4

(34) Sample B: Li.sub.1.00Ni.sub.0.41Fe.sub.0.11Ti.sub.0.02Mn.sub.1.46O.sub.4

(35) Sample C: Li.sub.1.02Ni.sub.0.48Fe.sub.0.11T.sub.0.02Mn.sub.1.37O.sub.4

Samples D and E According to the Present Invention

(36) Sample D: Li.sub.1.00Ni.sub.0.51Fe.sub.0.11Ti.sub.0.02Mn.sub.1.36O.sub.3.84F.sub.0.16

(37) Sample E: Li.sub.1.02Ni.sub.0.51Fe.sub.0.11Ti.sub.0.02Mn.sub.1.35O.sub.3.73F.sub.0.27

(38) All samples were doped with titanium. Samples B to E additionally comprised Fe, while samples D and E additionally contained fluorine, so that only these two samples D and E fall within the class of materials according to the present invention. Samples A to C should therefore be regarded as comparative samples.

(39) The composition of the samples was determined by means of chemical analysis and corresponds to the stoichiometric amounts weighed in. Except for sample A, which contained a small proportion of Li.sub.xNi.sub.1-xO as a foreign phase, all samples were phase-pure according to X-ray diffraction. All samples were disordered spinels which, according to BET measurements, had specific surface areas of from 2.9 to 5.9 m.sup.2/g and, according to scanning electron micrographs, had comparable morphologies. Since the synthesis conditions were chosen to be the same for all samples, the differences in the electrochemical properties shown in FIGS. 1-7 actually show the respective influences of doping.

(40) For the cell tests, 2-electrode Swagelok cells having lithium as the anode, 1M LiPF.sub.6 in EC/DMC 1:1 as the electrolyte and the samples A to E as the active material of the cathode were used. The cathode consisted of 80% of active material, 10% of polyvinylidene fluoride (PVDF), and 10% of carbon black. Specific values indicated are based on the mass of the active material.

(41) FIG. 1 shows the specific discharging capacities of samples A to E at 23° C. and different voltage ranges. In the high-voltage range, i.e., in the range from 3.5 V to 5 V, stable cycling was possible for all pure-phase samples. The advantages of the class of the materials according to the present invention in samples D and E become apparent, however, over the total voltage range. Over the first 50 cycles, there is, as shown in FIG. 1, no decrease in the specific capacity to be observed for these two samples. Sample D has a specific capacity of 176 mAh/g after 10 cycles.

(42) If the theoretical specific capacity of 147 mAh/g corresponds to the lithium intercalation of one lithium into Li.sub.1-wNi.sub.0.5Mn.sub.1.5O.sub.4 where 0≦w≦1 in the high-voltage range, more than one lithium per formula unit is, as shown in FIG. 2, cycled in the voltage range from 5.0 V to 2.0 V in the case of samples B to E after 10 cycles, since these have a specific capacity higher than 147 mAh/g.

(43) As shown in FIG. 3, in contrast, only the samples D and E according to the present invention display comparably high specific capacities as after 10 cycles after 100 cycles. Sample D is the only sample which still displays a specific capacity of 174 mAh/g, i.e., greater than 147 mAh/g, after 100 cycles. It is thus demonstrated that significantly more than one lithium per formula unit is intercalated in the case of sample D over 100 cycles at a C rate of C/2.

(44) FIG. 4 shows the coulombic efficiency of the samples in the first 30 cycles; this is defined as the ratio of discharging capacity to charging capacity. A coulombic efficiency of less than 1.0 indicates irreversible electrochemical secondary reactions. Possible causes are SEI formation, electrolyte decomposition, or dissolution of Mn.sup.2+ in the electrolyte. It is conspicuous in FIG. 4 that the coulombic efficiency is less than 1 for all samples. It is obvious that this reduction in the coulombic efficiency occurring even in the high-voltage range is attributable to the electrolyte used not being stable at above 4.5 V. It is interesting, however, that in the case of the two samples D and E which have been fluorine-doped according to the present invention there is, in contrast to samples A to C, no significant further reduction in the coulombic efficiency in the entire voltage range of 2.0-5.0 V, which represents an improvement over the samples which have not been fluorine-doped.

(45) FIG. 5 shows the cycle-dependent specific discharging energy and FIG. 6 shows the efficiency of the samples. As in the case of the specific capacity, the samples D and E also cycle significantly more stably in terms of the energy density over the entire voltage range from 5.0 V to 2.0 V than the samples A to C which were not doped with fluorine. In the case of sample D, the energy density decreases slightly after 50 cycles in the right-hand graph. This is related to a voltage hysteresis in the 3V range which increases with increasing number of cycles, as the comparison of FIG. 2 and FIG. 3 for this sample shows.

(46) The efficiency comprises energy losses due to the coulombic efficiency and losses due to the voltage difference between charging and discharging. Since cycling was carried out at a charging and discharging rate of C/2, i.e., 2 hours charging and 2 hours discharging, the efficiency shown includes kinetic losses which lead to an overvoltage. The decrease in the efficiency is conspicuous in a cycle over the entire voltage range. The reason lies in the greater voltage hysteresis in the 3V range, which is caused by the energy losses in the phase transformation from cubic to tetragonal. It is conspicuous that improved efficiencies are once again achieved by the samples fluorine-doped according to the present invention.

(47) FIG. 7 shows the specific capacities at 45° C. Samples D and E can be cycled stably over the first 50 cycles in the high-voltage range, which represents an improvement compared to the high-voltage spinels which have not been fluorine-doped in samples A to C. Relative improvements are likewise apparent over the entire voltage range. It is expected that the cycling stability can also be increased further at elevated temperature both by optimization of the material according to the present invention and by optimization in cell manufacture.

(48) The present invention is not limited to embodiments described herein; reference should be had to the appended claims.