Tungsten-substituted titanium-niobium mixed oxide active material

Abstract

An active material intended for the manufacture of an electrode, the active material comprising a substituted monoclinic titanium-niobium mixed oxide capable of allowing the insertion and extraction of Li+ ions, the active material having the following empirical formula (I): Ti(1+x)Nb(2-2x)WxO7 (I) wherein x is selected from the range of from 0.1 to 0.2.

Description

[0001] DESCRIPTION

[0002] TITLE: Active material made of tungsten-substituted niobium titanium mixed oxide

[0003] The present invention relates to the field of active materials intended to form an electrode for lithium accumulators. In particular, the invention relates to an active material formed from particles of mixed niobium and titanium oxide, part of the niobium of which is substituted by tungsten and titanium. According to a second aspect, the invention provides a method for manufacturing such an active material. According to other aspects, the invention provides an electrode formed from said active material and an electrochemical generator, in particular of the battery type, which comprises a negative electrode in said active material.

[0004] Lithium batteries are increasingly used as stand-alone power sources, particularly in portable equipment, where they have replaced nickel-cadmium (Ni-Cd) and nickel-metal hydride (Ni-MH) batteries, and now also for electric mobility. This development is explained by the continuous improvement in the performance of lithium batteries combined with a drastic reduction in their production costs, thus giving them mass and volume energy densities significantly higher than those offered by the Ni-Cd and Ni-MH sectors. While the first Li-ion batteries had an energy density of around 85 Wh / kg, more than 200 Wh / kg can now be obtained (energy density relative to the mass of the complete Li-ion cell).For comparison, Ni-MH batteries have a maximum of 100-110 Wh / kg and Ni-Cd batteries have an energy density of around 50-70 Wh / kg. This, combined with falling costs, explains why lithium batteries are now the most sold. New generations of more efficient lithium batteries are being developed for increasingly diverse applications (hybrid or all-electric cars, photovoltaic cell energy storage, etc.). In order to meet the ever-increasing demands for energy and sometimes power (per unit of mass and / or volume), new, even more efficient Li-ion battery electrode materials are essential.

[0005] The active electrode compounds used in commercial accumulators are, for the positive electrode, lamellar compounds such as LiCoC>2, LiNiOa and the mixed Li(Ni, Co, Mn, AI)C>2 OR spinel structure compounds of composition close to LiMn2O4. The negative electrode is generally carbon (graphite, coke, etc.) or possibly spinel Li4TisOi2 or a metal forming an alloy with lithium (Sn, Si, etc.). The theoretical and practical specific capacities of the negative electrode compounds mentioned are respectively approximately 370 mAh / g for graphite and 170 mAh / g for titanium oxide.

[0006] Despite its low capacity, compared to graphite, the Li4TisOi 2 compound finds its place on the market thanks to its high working potential, approximately 1.6V vs Li+ / Li, which makes it very safe and thanks to very good cyclability at high speed, which makes it the negative electrode material of choice for power applications.

[0007] Many studies have been carried out to find a compound with the same advantages as Li4TisOi2 in terms of both power and safety and having a higher specific capacity. Thus, niobium oxides and mixed Ti-Nb oxides were considered, niobium having a working potential close to that of titanium and allowing the exchange of 2 electrons per metal atom (Nb 5+ / Nb 3+ ).

[0008] The most interesting compounds are the oxides TiNb2O7 and Ti2NbioC>29. They have very high theoretical capacities (388mAh / g and 396mAh / g, respectively) compared to Li4TisOi2 (175mAh / g) and have a working potential close to that of Li4Ti50i2, which allows them to retain the latter's safety advantages. They are therefore very interesting candidates for replacing the latter for applications requiring more energy.

[0009] However, the TiNb2O7 material, which is more attractive in terms of cost due to the higher Ti / Nb ratio than in the Ti2NbioC>29 material, has limitations in terms of power performance and cyclability.

[0010] The present invention aims to overcome the drawbacks mentioned above. To this end, the present invention provides an active material for the manufacture of an electrode, the active material comprising a monoclinic mixed oxide of substituted niobium titanium, capable of allowing the insertion and extraction of Li-i- ions, the active material having the following empirical formula (I): Ti(1 + x)Nb(2-2x)WxO7 (I) in which the value x is chosen in the range from 0.05 to 0.2.

[0011] This active material requires the substitution of a portion of the niobium in the mixed oxide TiNb2O7 with tungsten and titanium while preserving the initial crystalline structure. As will be seen later in Figure 2, only the volume of the mesh increases with the tungsten content. The niobium content decreases at the same time as the tungsten content increases, as does the titanium content to maintain the initial stoichiometry.

[0012] Thus optimized, this new active material offers stabilization of cycling performance over time while maintaining high working potential and limiting performance losses during slow initial cycling.

[0013] According to one arrangement, the active material consists of a single monoclinic mixed oxide of substituted niobium titanium of the following empirical formula (I):

[0014] Ti(1 +x)Nb(2-2x)WxO7 (I) in which the value x is chosen in the range from 0.05 to 0.2.

[0015] According to one possibility, the value x is chosen in the range from 0.10 to 0.2, in particular in the range from 0.12 to 0.18, and for example x = 0.15. In the active material according to the invention, titanium has an oxidation state + IV and niobium has an oxidation state +V. The oxidation states of the metals in Ti(i +X )Nb(2-2x)WxO7 are identical to those of the unsubstituted compound TiNb2O7. This is advantageous because a reduction in the oxidation states, from Ti4+ to Ti3+ or from Nb5+ to Nb4+ would de facto limit the number of electrons available during the first lithiation (metal reduction) and therefore reduce the capacity of the material.

[0016] Concretely, the active material may comprise particles having an average diameter Di greater than 100 nm and less than or equal to 0.5 mm. These average diameter Di values ​​meet the density / compactness requirements of the active material to provide a satisfactory energy density. The particles have a slightly elongated spherical overall shape and whose surface has contours with a certain irregularity, the average diameter Di of the particles corresponds to the average value of the three dimensions measured by laser granulometry.

[0017] According to one arrangement, the active material consists solely of said particles.

[0018] According to one possibility, the active material consists of a mixed oxide of Ti(1 +x)Nb(2-2x)WxO7.

[0019] According to one arrangement, the active material is intended for the manufacture of an electrode for Li-ion accumulators. The use of this active material makes it possible to achieve greater cycling stability and to limit the loss of performance observed with x = 0.

[0020] In particular, the particles can be divided into three populations of average diameters Di, a first population having an average diameter D1 with 0.1 pm < D1 < 0.8 pm, a second population having an average diameter D2 with 1 pm < D2 < 10 pm, and a third population having an average diameter D3 with 10 pm < D3 < 0.5 mm.

[0021] According to another embodiment, the particles are divided into two populations of average diameters Di, a first population having an average diameter D1 with D1 < 0.8 pm, and a second population having an average diameter D2 with D2 > 1 pm.

[0022] According to one arrangement, the first population of particles has an average diameter D1 with 0.1 micrometer < D1 < 0.8 pm

[0023] Alternatively, the second particle population has a mean diameter D2 with 1 pm < D2 < 0.5 mm.

[0024] According to a second aspect, the invention provides a method for manufacturing the active material as previously described, which comprises solid-state synthesis.

[0025] According to one arrangement, solid-state synthesis is intended to lead to particles of active material. According to one possibility, solid-state synthesis is carried out from precursor reagents, in particular solid precursor reagents.

[0026] Concretely, the precursor reagents are TiOa, Nb20s and WO3.

[0027] According to one feature, the precursor reagents are used in stoichiometric proportions.

[0028] According to one provision, the method comprises the steps of:

[0029] - a) grinding of the precursor reagents in powder form in a planetary ball mill so as to obtain a homogeneous powdery mixture,

[0030] ■ b) calcination by applying a heat treatment at a temperature between 900°C and 1200°C so as to obtain the active material.

[0031] According to one arrangement, the method comprises, after step b), carrying out a step c) of low-energy grinding of the active material so as to reduce any agglomerates and obtain a homogeneous powder having particles with an average diameter Di greater than 100 nm and less than or equal to 0.5 mm.

[0032] One possibility of low-energy grinding is manual grinding, such as using a mortar and pestle.

[0033] Concretely, step a) of grinding the precursor reagents can be carried out at a speed of approximately 400 rpm, it notably includes an alternation of grinding sequences and rest sequences.

[0034] According to one embodiment, solid-state synthesis leads to particles without agglomerate, for example having a diameter greater than 100 nm.

[0035] This helps to partially meet the density / compactness requirements needed to achieve energy densities satisfactory for the intended objective.

[0036] According to another aspect, the invention provides an electrode comprising the active material as previously described.

[0037] According to yet another aspect, the invention provides an electrochemical generator, in particular of the battery type, which comprises a positive electrode and a negative electrode comprising the active material as previously described and a non-aqueous electrolyte comprising lithium.

[0038] The active materials proposed by the invention can be adapted to high power requirements (fast charge / fast discharge, associated with very good cyclability), and maintaining an energy density at a high level (>100Wh / kg).

[0039] Other characteristics and advantages will appear on reading the detailed description below, of a non-limiting example of implementation, made with reference to the appended figures in which:

[0040] [Fig. 1] represents a laser granulometric analysis of different compositions of active materials according to an embodiment of the invention. [Fig. 2] represents an X-ray diffraction (XRD) diagram of the different compositions of active materials.

[0041] [Fig. 3] represents a graph illustrating the evolution of the charged capacity as a function of the number of cycles of button batteries obtained from the different compositions of active materials.

[0042] [Fig. 4] shows a graph illustrating the specific load capacity as a function of the number of cycles.

[0043] With reference to the gross formula (I) illustrated below, the present invention provides an active material intended for the manufacture of electrodes for lithium accumulators based on a mixed niobium titanium oxide substituted by tungsten:

[0044] Ti(1 +x)Nb(2-2x)WxO7 (I) in which the value x is chosen in the range from 0.05 to 0.25.

[0045] This active material is obtained by solid-state synthesis from a mixture of stoichiometric proportions of powders of the precursor reagents comprising TiC>2, Nb2O5 and WO3 (refer to Table 1). TiO2 is for example obtained from Huntsmann (TiO2anatase purity 99.0%), Nb20s is obtained from Sigma Aldrich (purity 99.9%) and WO3 is obtained from Sigma Aldrich (purity 99.9%). The initial oxidation states of Ti +IV and Niobium +V are retained during the reaction.

[0046] Table 1: Masses of precursor reagents used for Ti synthesis<i+x> Nb(2-2x)WxO?

[0047] The process for manufacturing the active material comprises a step a) of grinding the precursor reagents present in powder form in a planetary ball mill (PM 100 CM, Retsch - 32 g of agate balls in a bowl also made of agate for a mass of precursor reagents of approximately 7 g) with a speed of 400 rpm for 8 hours. The grinding is not carried out continuously but by repeating the grinding sequence and the rest sequence, each lasting 5 min.

[0048] Once the powdery mixture is perfectly homogeneous, it is placed in three alumina crucibles (diameter of 3 cm each) then placed in a muffle furnace (Carbolite, Model CWF 1200) until it reaches a temperature of approximately 1,100°C for 16 hours for example (step b). At the end of the heat treatment, the active material is obtained in particulate form.

[0049] The recovered active material is ground manually in an agate mortar. Five minutes of grinding are sufficient to obtain a homogeneous powder in the form of particles with an average diameter Di greater than 100 nm and less than or equal to 0.5 mm. As shown in Figure 1 illustrating the laser particle size distribution diagram (left column), the average diameter Di is divided into three populations similar to those of the tungsten-free niobium titanium mixed oxide (MALVERN MASTERSIZER device used). These three populations include a first population with an average diameter D1 with 0.4 micrometers < D1 < 0.8 micrometers, a second population with an average diameter D2 with 1 micrometer < D2 < 10 micrometers, and a third population with an average diameter D3 with 10 micrometers < D3 < 0.5 mm.

[0050] According to one possibility, the particles are divided into two populations of average diameters Di, a first population having an average diameter D1 with D1 < 0.8 pm, and a second population having an average diameter D2 with D2 > 1 pm.

[0051] Also illustrated in Figure 1 (right column), the active material particles subjected to ultrasonic treatment lead to the same particle size distribution, which shows that the particles obtained at the end of step c) are free of agglomerate.

[0052] Illustrated in Figure 2, an X-ray diffraction analysis of the particles obtained at the end of step c) makes it possible to verify that the crystalline structure of Ti2NbsO? is identical to that of the active material Ti(i +X)Nb(2-2x)WxO7 for all W substitution rates (x from 0.05 to 0.2). The same analysis performed on the mixed oxide of niobium and titanium substituted with the value x=0.25 (not shown) shows that impurities are present within the active material, notably TiO2. It would seem that this degree of W doping marks the limits of the integration of tungsten in the structure of the mixed oxide.

[0053] Electrodes and button cells were designed by traditional means in order to observe the properties obtained by the active material according to the different values ​​of x and Ti2NbsC>7.

[0054] The active material Ti(i+x)Nb(2-2x)WxC>7 (with x between 0.05 and 0.25) and a carbon additive (Carbon Black SUPER C65 from TIMCAL) are mixed and ground manually in an agate mortar in cyclohexane (Purity >99.5%, Merck ENSURE) for 5 minutes.

[0055] After complete evaporation of the solvent, a solution of polyvinylidene fluoride (PVDF, Solef 5130, Solvay) at 8% by mass in N-methyl-2-pyrrolidone (NMP, Purity >99%, Merck) is added to the ground mixture until a mass formulation of 80% active material, 10% electronic conductor and 10% binder for a dry extract of 30% is reached. Table 2 illustrates the proportions of the reagents used. Table 2: Typical masses for the preparation of Ti(i+x)Nb(2-2x)W inks x O7

[0056] The resulting ink is then coated onto an aluminum foil using a doctor blade with a thickness of 100 μm. After 24 hours of drying at 55°C, electrodes with a diameter of 14 mm are cut and passed through a 10-ton press.

[0057] The electrodes are then assembled in a glove box to form CR2032 button cells. The counter electrode is made of lithium metal, one of the separators is made of polypropylene felt (Viledon, Freudenberg) and the other separator is made of polypropylene (CG2500, Celgard). The electrolyte consists of a mixture of ethylene carbonate (EC) / propylene carbonate (PC) / dimethyl carbonate (DMC) (1:1:3 vol) with lithium hexafluorophosphate (LiPF6) (1 M) (LP100, UBE Industries).

[0058] The performances of the button cells obtained are gathered in figure 3 which illustrates, on the left side, the evolution of the capacity in Lithiation of the Ti(i +X)Nb(2-2x)WxO7 with x = 0; 0.05; 0.1; 0.15 and 0.2 depending on the number of cycles and at different regimes (respectively cycling at C / 10, C, 2C, 3C, 5C and 10C). The right part of Figure 3 illustrates the evolution of the Lithiation capacity of the same compounds with cycling at the C / 10 regime from 31 e cycle.

[0059] As can be observed, the poorest performances are obtained when the electrode is TiNb2C>7 for which the loss is quickly very significant. The cycling stability and the performance losses during the cycles are better with the substituted active material Ti(i+x)Nb(2-2x)W xO7. Also shown in Table 3, which reports the degradation values ​​at C / 10 for each of the active materials, the best results are obtained with x = 0.1 and x = 0.15. The performance losses vary from single to double after 100 cycles between an active material in which x = 0 (loss of approximately 30%) and an active material in which x = 0.1 and 0.15 (loss of 16.7 and 14.1% respectively). The improvement in lithiation capacities is also the best for x = 0.1 and 0.15.

[0060] Figure 4 illustrates the performance of a button cell obtained with x = 0.25, i.e. the cycling stability and the performance losses during the cycles for x = 0.25 (on the abscissa the number of cycles and on the ordinate the specific capacity values ​​in mAh / g). At the date of filing of this application, the analyses have not yet been completed. Nevertheless, it is already possible to see that the initial capacity at C / 10 of the button cell for x = 0.25 is lower than those obtained with values ​​of x between 0.10 and 0.20.

[0061] The power capacity (on the left respectively with cycles at C / 10; C, 2C, 3C, 5C and 10C for ease of comparison) is 120mAh / g at 10C. Although correct, this capacity remains lower than the optimum obtained which is greater than 130.

[0062] Finally, the cycling performance is poor: after 37 cycles, the specific capacity is 207 mAh / g (right part of Figure 4), which is lower than the values ​​obtained at 40 cycles for batteries in which x is between 0.10 and 0.20 (more than 230 mAh / g - Figure 3). The performance degradation at cycle 40 is 10.2, which is already equivalent to the performance loss for cycle 60 with x = 0.2 (refer to Table 3). An estimate made for cycle 60 indicates a degradation of 24.8%. These results seem to corroborate the fact that the integration of tungsten into the mixed oxide of niobium and titanium would reach limits around x = 0.25.

[0063] Table 3: Performance loss compared to initial capacity at C / 10 (cycle X. Vs cycle 31)

[0064] Thus, the present invention proposes, according to one embodiment, an active material consisting of a mixed oxide of Ti(i +X)Nb(2-2x)WxO7 intended for the manufacture of an electrode for Li-ion accumulators. The use of this active material makes it possible to achieve greater cycling stability and to limit the loss of performance observed with x = 0. As visible in Figure 3 and Table 3, the optimum is obtained for a tungsten substitution with a value of x = 0.15. The manufacturing process of the material comprises a solid-state synthesis, leading for example to agglomerate-free particles, which may have a diameter greater than 100 nm, which can partially meet the density / compactness requirements necessary to achieve satisfactory energy densities for the intended purpose. The active materials proposed by the invention are thus adapted to high power requirements (fast charging / fast discharging, associated with very good cyclability), and maintaining an energy density at a high level (>100Wh / kg).

[0065] It goes without saying that the invention is not limited to the embodiment variants described above by way of example but that it includes all technical equivalents and variants of the means described as well as their combinations.

Claims

CLAIMS 1. Active material for the manufacture of an electrode, the active material comprising a monoclinic mixed oxide of substituted niobium titanium, capable of allowing the insertion and extraction of Li+ ions, the active material having the following empirical formula (I): Ti(1 + x)Nb(2-2x)WxO7 (I) where x is chosen from the range 0.10 to 0.

20.

2. Active material according to claim 1, which comprises particles divided into two populations of average diameters Di, a first population having an average diameter D1 with D1 < 0.8 pm, a second population having an average diameter D2 with D2 > 1 pm.

3. Active material according to claim 1 or 2, which comprises particles having an average diameter Di greater than 100 nm and less than or equal to 0.5 mm.

4. Method for manufacturing the active material according to one of claims 1 to 3, which comprises solid-state synthesis.

5. Manufacturing method according to claim 4, wherein the precursor reagents are TiC, Nb20s and WO3.

6. Manufacturing method according to claim 4 or 5, in which the precursor reagents are used in stoichiometric proportions.

7. Manufacturing method according to one of claims 4 to 6, the method comprising the steps of: - a) grinding of the precursor reagents in powder form in a planetary ball mill so as to obtain a homogeneous powdery mixture, ■ b) calcination by applying a heat treatment at a temperature between 900°C and 1200°C so as to obtain the active material.

8. Method for manufacturing the active material according to claim 7, which comprises, after step b), carrying out a step c) of low-energy grinding of the active material so as to reduce any agglomerates and obtain a homogeneous powder having particles with an average diameter Di greater than 100 nm and less than or equal to 0.5 mm.

9. Electrode comprising the active material according to one of claims 1 to 3.

10. Electrochemical generator, in particular of the battery type, which comprises a positive electrode and a negative electrode comprising the active material according to one of claims 1 to 3 and a non-aqueous electrolyte comprising lithium.

Images