Method for Manufacturing Mixed Oxide Powders as Well as a Mixed Oxide Powder

Abstract

A method for manufacturing mixed oxide powders including the steps (a) producing a raw material mixture, (b) bringing the raw material mixture into a hot gas flow for the thermal treatment in a reactor, (c) forming particles of the mixed oxide powder, and (d) bringing the particles of the mixed oxide powder which are obtained in the step (b) and (c) out of the reactor, wherein the raw material mixture is manufactured in the form of a solution or suspension of at least one salt and/or salt mixture of at least one compound of the elements lithium, nickel and/or manganese, as well as a mixed oxide powder which is manufactured according to this method.

Claims

1. A method for manufacturing mixed oxide powders comprising the steps of: (a) producing a raw material mixture, (b) bringing the raw material mixture into a hot gas flow for a thermal treatment in a reactor, (c) forming particles of the mixed oxide powder, and (d) bringing the particles of the mixed oxide powder which are obtained in the steps (b) and (c) out of the reactor. wherein the raw material mixture is manufactured in the form of a solution or dispersion, wherein the raw material mixture comprises at least one of the elements lithium, nickel and/or manganese.

2. The method according to claim 1, wherein the raw material mixture comprises at least two of the elements lithium, nickel and/or manganese.

3. The method according to claim 1, wherein one or more dopants selected from the elements magnesium, aluminium, titanium, vanadium, chromium, iron, cobalt, copper, zinc, silicon, zirconium, ruthenium, rhodium, palladium, silver, and platinum are added to the raw material mixture in step (a).

4. The method according to claim 1, wherein a stoichiometric raw material mixture is produced in step (a).

5. The method according to claim 1, wherein a single-stage or multi-stage wet-chemical intermediate step is carried out in the reactor before the thermal treatment.

6. The method according to claim 1, wherein the method for manufacturing mixed oxide powders is spray pyrolysis.

7. The method according to claim 1, wherein the raw material mixture is introduced into a pulsating hot gas flow for the thermal treatment in the reactor.

8. The method according to claim 7, wherein a pressure amplitude and an oscillation frequency of the pulsating hot gas flow can be set independently of one another.

9. The method according to claim 1, wherein the hot gas flow for the thermal treatment of the raw material mixture in the reactor has temperatures between 200° C. and 2500° C.

10. The method according to claim 1, wherein a cooling gas is fed to the hot gas flow before step (d).

11. The method according to claim 1, wherein the particles are separated from the hot gas flow for bringing the particles which are obtained in steps (b) and (c) out of the reactor.

12. The method according to claim 11, wherein the particles are separated from the hot gas flow at temperatures above 200° C.

13. The method according to claim 1, wherein the particles which are obtained from the reactor are subjected to a post-treatment, wherein the post-treatment is at least a grinding and/or a thermal post-treatment.

14. The method according to claim 13, wherein the particles which are obtained from the reactor are firstly subjected at least to the grinding and subsequently at least to the thermal post-treatment.

15. A mixed oxide powder in the form of LiNi.sub.yMn.sub.2-yO.sub.4 particles manufactured in a hot gas flow from a raw material mixture in the form of a solution or dispersion, wherein the raw material mixture comprises at least one of the elements lithium, nickel and/or manganese.

16. The mixed oxide powder according to claim 15, wherein the mixed oxide powder is manufactured in a hot gas flow from a raw material mixture in the form of a solution or dispersion, wherein the raw material mixture comprises at least two of the elements lithium, nickel and/or manganese.

17. The mixed oxide powder according to claim 15, wherein the raw material mixture in the form of a solution or dispersion is manufactured from at least one salt and/or salt mixture.

18. The mixed oxide powder according to claim 15, wherein the mixed oxide powder comprises one or more dopants selected from the elements magnesium, aluminium, titanium, vanadium, chromium, iron, cobalt, copper, zinc, silicon, zirconium, ruthenium, rhodium, palladium, silver, and platinum.

19. The mixed oxide powder according to claim 15, wherein the mixed oxide powder has an electrochemical capacity of larger than 100 mAh/g at 0.1 C.

20. The mixed oxide powder according to claim 15, wherein the mixed oxide powders have a particle size of 0.5 μm to 100 μm.

21. The mixed oxide powder according to claim 15, wherein the mixed oxide powders have a cubic crystal system.

22. The mixed oxide powder according to claim 15, wherein the mixed oxide powder is manufactured according to a method for manufacturing mixed oxide powders comprising the steps of: (a) producing a raw material mixture, (b) bringing the raw material mixture into a hot gas flow for thermal treatment in a reactor, (c) forming particles of the mixed oxide powder, and (d) bringing the particles of the mixed oxide powder which are obtained in the steps (b) and (c) out of the reactor.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The invention is hereinafter explained in more detail by way of the attached drawings. In these are shown in:

[0029] FIG. 1 a powder diffraction diagram (XRD diagram) of all LiNi0.5Mn1.5O4 particles in a 2θ-region of 10° to90°,

[0030] FIG. 2 a powder diffraction diagram (XRD diagram) of all LiNi0.5Mn1.5O4 particles in a 2θ region of 35° to 70° with LixNi1-xO phase impurities which are highlighted in black boxes,

[0031] FIG. 3 FESEM pictures for the different LiNi0.5Mn1.5O4 particles,

[0032] FIG. 4 the particle size distribution of the different LiNi0.5Mn1.5O4 particles,

[0033] FIG. 5 an overview over the specific surface area and the particle size distribution (d90) of LiNi0.5Mn1.5O4 particles,

[0034] FIG. 6 rate capability and ageing test results for LiNi0.5Mn1.5O4 particles,

[0035] FIG. 7 an overview over the specific surface area, correlated with the specific discharge capacity at a 0.1 C-rate, and particle size distribution (d90) of LiNi0.5Mn1.5O4 particles,

[0036] FIG. 8 discharge curves of LiNi0.5Mn1.5O4 particles at 0.1 C,

[0037] FIG. 9 cumulative end point capacities of LiNi0.5Mn1.5O4 particles, (charging: full symbol; discharging: empty symbol) which are measured during galvanostatic cyclisation,

[0038] FIG. 10 a correlation between specific surface area and cumulative discharge of LiNi0.5Mn1.5O4 particles,

[0039] FIG. 11 an impedance measurement of LiNi0.5Mn1.5O4 particles given a state of charge (SOC) of 90%,

[0040] FIG. 12 an equivalent circuit diagram for the electrochemical impedance spectroscopy (EIS) adaptation,

[0041] FIG. 13 values for the surface film resistance (Rsf) plotted against the state of charge (SOC), said values having been determined by analysis of the EIS measurements,

[0042] FIG. 14 values for the charge transfer resistance (Rct) plotted against the state of charge (SOC), said values having been determined by analysis of the EIS measurements;

[0043] FIG. 15 values for the electrolyte resistance (Re) plotted against the state of charge (SOC), said values having been determined by analysis of the EIS measurements,

[0044] FIG. 16 a correlation between the specific surface area space and surface film resistance (Rsf) of the LiNi0.5Mn1.5O4 particles.

DETAILED DESCRIPTION

[0045] On account of the high potential versus lithium, spinel LiNi0.5Mn1.5O4 (LNMO) is a promising candidate amongst other high-potential materials, in order to fulfil the demands of technological progress, in particular in order to achieve a higher energy and power density. LNMO has a dominant potential plateau at 4.7 V versus Li/Li.sup.+ and the theoretical specific capacity of 147 mAh/g. Furthermore, LNMO can significantly reduce the raw material costs which occur on manufacture and has a reduced toxicity compared to cathode materials based on cobalt. The spinel structure exhibits a high structural stability due to isotropic expansion and contraction during intercalation. Depending on the distribution of the nickel and manganese atoms, LNMO can form two space groups with an ordered and unordered crystal structure. In the ordered structure (space group of the primitive cubic structure P4332), Ni and Mn atoms have the position 4a and 12d, whereas in the unordered structure (space group of the surface-centred cubic structure Fd3m) both types of atoms are randomly distributed onto the octahedral positions 16d.

[0046] The electrochemical performance of this cathode material is predominantly dependent on two parameters, the presence of LiyNi1-yO rock-salt phase (foreign phase) and the quantity of Mn.sup.3+ ions. It is difficult to synthesise a pure LNMO phase with a controlled morphology due to the undesired particle growth of foreign phases during the sintering process at high temperatures. The formation of the unordered LNMO is usually accompanied by the formation of a rock salt phase which however reduces the achievable capacity of the cathode material. For this reason, it is important to reduce the formation of the rock salt phase to a minimum, so that the manufactured LNMO particles have little as possible contamination phases.

[0047] A further challenge is the control of the content of Mn.sup.3+ ions in the crystal. If for example the quantity of Mn.sup.3+ ions in the crystal is too high and two Mn.sup.3+ ions interact, an Mn.sup.2+ ion and a redox-inactive Mn.sup.4+ ion form. This reaction is denoted as a disproportionisation reaction. Furthermore, the Mn.sup.2+ ion is soluble in the electrolyte and the Mn.sup.4+ ion is stable and electrochemically active. This leads to a poor performance of ordered LNMO and to a loss in capacity. The conductivity of the unordered LNMO is amplified by small quantities of Mn.sup.3+ ions and hence exhibits an improved rate capability. The diffusion of lithium ions depends on the composition and morphology of the LNMO. The synthesis of the specific space groups of LNMO depends on the calcination temperature. The calcination temperature must lie at about 700° C. in order to obtain the ordered spinel structure, whereas temperatures of 700° C. and more are necessary for unordered spinel structures.

[0048] In the series of trials which were carried out, the unordered LNMO crystal structure was synthesised, since in literature it has been found that it has improved electrochemical characteristics and an improved stability compared to the ordered crystal structure.

[0049] In the series of trials, mixed oxide powder in the form of LNMO with an unordered crystal structure was synthesised by way of different manufacturing methods. The first synthesis method is based on a spray drying method, the second synthesis method on a manufacturing method similar to spray pyrolysis. In both manufacturing methods, the LNMO particles are produced in a hot gas flow which flows through the reactor and are separated out of this after synthesis, for example by way of a separating device. Furthermore, the LNMO powder which was manufactured by the two synthesis methods was subjected to a post-treatment. As a post-treatment, on the one hand a single-stage thermal post-treatment (1) and on the other hand a two-stage mechanical and thermal post-treatment (2) were carried out:

[0050] With the single-stage thermal post-treatment, the precursor was firstly heated at 1K/min to 200° C., followed by 5 K/min to800° C. and subsequently treated further at a temperature of 800° C. for 5 hours amid air in a muffle furnace. After the five-hour treatment time, the mixed oxide powder in the form of LNMO particles was cooled overnight in the furnace.

[0051] In the case of the two-stage mechanical treatment, firstly a grinding step with a grinding of the LNMO particles for 1 hour in a planet ball mill (company FRITSCH GmbH) with EtOH as a grinding medium and subsequently a thermal post treatment in a muffle furnace were carried out. With the thermal post-treatment, the LNMO particles were heated at 1 K/min to 200° C. followed by 5 K/min to 800° C. and subsequently treated further for duration of 5 hours. The mixed oxide powder was cooled overnight in the furnace after the treatment time.

[0052] Table 1 shows the mixed oxide powder, in particular LiNi0.5Mn1.5O4 particles (LNMO) which is manufactured according to a method according to the invention. Hereby, the particles which are manufactured by way of the pulsating gas flow are characterised by the abbreviation AT, and the particles which are manufactured by way of the spray drying by the abbreviation ST. The particles which are manufactured by way of the two aforementioned method and which have been synthesised on the basis of a suspension are marked by S and those on the basis of a solution are marked by L. If the synthesised particles have undergone no additional temperature treatment, then the particles are characterised by 0, in the case of an additional 5-hour temperature treatment are characterised by 5 and in the case of an additional 5-hour temperature treatment with a prior 1-hour grinding step by 5g.

TABLE-US-00001 TABLE 1 examined LNMO particles which are synthesized according to different manufacturing methods APPtec APPtec Spray (solution) (suspension) drying without post-treatment AT_L_0 AT_S_0 ST_L_0 temperature post-treatment AT_L_5 AT_S_5 ST_L_5 (5 h at 800° C.) grinding (1 h) and AT_L_5g AT_S_5g ST_L_5g temperature post-treatment (5 h at 800° C.)

[0053] The APPtec method is a method which has been developed by the company Glatt Ingenieurtechnik, with which precursors are sprayed into a pulsating gas flow in different manners, for example on the basis of a solution or suspension, with an adjustable droplet size distribution. The pulsating gas flow generates special thermodynamic reaction conditions in a reactor, these giving the manufactured powders advantageous characteristics. The desired chemical and mineralogical reaction takes place by way of the defined thermal treatment, and particles form. The duration of the thermal treatment of the sprayed-in raw material mixtures is preferably less than 10 seconds.

[0054] A homogeneous precursor was manufactured for the synthesis of LNMO by way of spray drying and this was subsequently regulated in temperature for forming the suitable crystal phase. The raw material mixture was manufactured by way of dissolving stoichiometric quantities of CH3COOLi H2O (99%, Sigma-Aldrich), Ni(CH3COO)2 4 H2O (99%, Sigma-Aldrich) and Mn(CH3COO)2 4 H2O (99%, Sigma-Aldrich) in ethanol. The solution was sprayed into a Büchi Mini Spray Dryer B-290 for manufacturing the precursor.

[0055] The mixed oxide powders in the form of LMNO, developed with the aforementioned methods and possibly amid post-treatment were characterised and tested, in order to understand the influence of the different synthesis methods upon the electrochemical performance.

[0056] Undesirable reactions occur in the cell due to the high voltage and these represent a huge challenge for the commercialization of this high-voltage cathode material. These lead to a decomposition of the electrolyte, to a lower Coulomb efficiency and to an increase of the internal resistance over the complete cycle and hence influence the service life of the battery to a significant extent. The cumulative end point shift capacity and the electrochemical impedance spectroscopy (EIS) were examined, in order to measure and characterise the degree of undesirable reactions in the cells.

[0057] The performance analysis and the results of the cumulative specific capacity exhibit the fluctuations of the electrolyte break-down between the materials. EIS technology was used in order to examine the electrode materials, since it can describe the relation between the crystal lattice and the electrochemical characteristics.

[0058] The characterisation of the mixed oxide powder in the form of LNMO, manufactured by way of the different methods as described previously, was effected amid the aid of different analyses. The scanning electronic microscope (SEM) pictures were taken whilst using a Crossbeam NVISION 40 of Carl Zeiss SMT with an Everhart Thornley detector. X-ray diffraction (XRD) data was measured amid the use of a Bruker D8 diffractometer with a CuKα radiation (0.154056 nm). The electrodes for the electrochemical measurements were manufactured by way of pouring a slurry of 80% active material (LNMO), 10% binder (PVDF, Solef) and 10% carbon (Super P, Timical) onto Al foil (Hydro) and was dried for 20 hours at a temperature of 60° C. Cells with a two-electrode construction were provided as half-cells versus metallic lithium. A Whatmann glass fibre disc was used as a separator and LP40 (1M LiPF6 in EC:DEC 1:1 w/w BASF) was used as an electrolyte. The joined-together button cells LNMO/Li were galvanostatically cycled between 3.5 V and 5.0 V. All electrochemical measurements were carried out at 30° C. amid the use of a BASYTEC cell testing system. Five button cells were provided for each material, in order to obtain a reliable average. The button cells which were used for the galvanostatic measurement were likewise used for the EIS measurement. EIS was created with the Gamry-Framework Version 6.25, integrated with Basytec CTS at 30° C. The EIS measurements were carried out at

[0059] 10 mV interference amplitude in the range of 100 kHz to 1 Hz in automatic sweep mode from high to low frequencies. The counter-electrode is large enough so that it does not significantly influence the EIS behaviour of the operating electrode. The impedance was measured at a state of charge (SOC) of 90%, 60%, 30% and 10%. The potential was stabilised for 2 hours before the impedance was applied at each SOC. The produced impedance was further adapted amid the use of the function ZfitGUI (Varagin) of Matlab whilst using the equivalent circuit diagram which is shown in FIG. 12.

[0060] The results are represented and discussed hereinafter by way of diagrams.

[0061] XRD measurements were carried out in order to determine whether the Spinel phases were formed by the different synthesis methods. X-ray diffraction (XRD) is a unique method for examining the crystallinity of a substance. XRD is mainly applied for these questions:

[0062] Proof of identity of crystalline material (for official use or in development)

[0063] Proof of identity of different polymorphous shapes (“fingerprints”)

[0064] Differentiation between amorphous and crystalline material

[0065] Quantification of the percentage crystallinity of a particle.

[0066] The crystal imperfection of LNMO is related to the formation of oxygen vacancy sites which occurs when the particles are thermally treated at increased temperatures of more than 700° C. FIG. 1 shows the powder diffraction diagrams of all manufactured LNMO particles according to Table 1. The reflections of the crystal planes are highlighted in FIG. 1 and are characteristic of the LNMO crystal phases. It has been found that the particles have a different phase purity depending on their synthesis. The particles without an additional calcination either exhibit a low crystallisation, characterised by wide reflections with a low intensity for the LNMO particles which are manufactured in the pulsating hot gas flow (AT_L_0 and AT_S_0), or a large number of reflections, wherein these are not down to the reflections of the spinel phase which by way of their indices are specified for LNMO particles (ST_L_0) which are manufactured by way of spray drying. The weak reflections at 37.5°, 43.7° and 63.7° indicate impurities which are common and are also described in the literature and are marked by arrows and can be assigned to the rock salt phase LiyNi1-yO. These are often verified with the unordered LNMO crystal structure.

[0067] By way of applying an additional thermal treatment in the form of a calcination, it was possible to improve the crystallisation of the particles of the mixed oxide powder. This on the one hand is visible by way of the change of the solution-based LNMO particles which are manufactured in the pulsating hot gas flow, in comparison between untreated particles (AT_L_0) and treated particles (AT_L_5) and on the other hand by way of the reduction of the undesired phases for the other particle s (AT_S_5, ST_L_5) compared to the untreated versions (AT_S_0, ST_L_0).

[0068] For further improving the material characteristics, the LNMO particles were ground in a planetary ball mill for one hour before the second temperature treatment. The changes of the intensity curves with respect to the contamination phases are highlighted in FIG. 2 by boxes. It is shown that the additional treatment has a significant influence on some materials. In particular, as is illustrated in FIG. 2, the suspension-based particles (AT_S) which are manufactured in the pulsating hot gas flow exhibit a significant change in the intensity of the contamination phase. By way of the treatment, the contamination quantity for these particles was reduced (AT_S). A change for the other particles by way of the grinding was not ascertained. The reduction of the contamination phase was effected on account of the improved oxygen contact with the surface of the particles. The oxygen is necessary in order to integrate the rock salt phase back into the spinel structure. For this reason, the calcination together with the grinding improve the crystallisation once again.

[0069] It has been found that preferably more time is necessary in order for the crystals to be able to form and grow. Whereas the influence of the grinding on the solution-based particles (ST_L and AT_L) which only have low quantities of the rock salt phase is low, the additional grinding step exhibits a further improvement for particles with a high quantity of rock salt phase (AT_S_5). These phases are subjected to oxygen in an improved way and manner due to the grinding. FIG. 1 as a whole for the particles (*_5g) post-treated in a two-stage manner shows a low quantity of contamination phases.

[0070] FIG. 3 shows pictures of field emission scanning electron microscopy (FESEM) for all nine samples. The difference with regard to crystal morphology, crystal size and crystal make-up was examined by way of the FESEM pictures.

[0071] FIG. 3 shows that an additional calcination has a significant visible influence on the LNMO particles. Without thermal treatment, no crystal planes for the particles (AT_L, AT_S) manufactured in the pulsating hot gas flow can be recognised. In contrast to this, a few crystal planes are visible for the spray-dried LNMO particles (ST_L), wherein however a pure crystal is not to be found.

[0072] It is represented in FIG. 3 that crystals are formed by way of the single-stage post-treatment in the form of an additional calcination of the LNMO particles after the 5-hour thermal treatment. Furthermore, the materials exhibit different crystal shapes. The LNMO particles which are produced in the pulsating hot gas flow exhibit smaller crystals, and it is particularly the solution-based precursors which comprise (AT_L_5) crystals after an additional temperature treatment, wherein the particle size of the crystals is about 0.1 μm.

[0073] The particle size is around 1 μm for the suspension-based LNMO particles (AT_S_5) which are manufactured in the pulsating hot gas flow with a single-stage post-treatment, whereas even larger crystals with a particles size of greater than 1 μm are exhibited by the particles (ST_L_5) which are manufactured by way of spray drying given the same treatment. Pure crystal planes are recognisable after a single-stage post-treatment and the LNMO particles mostly form crystal in the form of octahedrons.

[0074] Furthermore, the FESEM pictures show no significant changes for the LNMO particles with a 2-stage post-treatment. The mechanical post-treatment in the form of a grinding step therefore only has little influence on the crystal size. The mechanical post-treatment has a greater influence on the aggregation and agglomeration if the LMNO particles.

[0075] As a whole, it is ascertained that an additional 5-hour temperature treatment is necessary in order to form crystals and assist in their growth. It is exclusively the suspension-based LNMO particles (AT_S) which are manufactured in the pulsating hot gas flow and LMNO particles (ST_L) which are manufactured by way of spray drying which have the preferred shape of spinel, the octahedron. This means that the metal ions distribute in a uniform manner during the temperature treatment, in order to form the preferred crystal planes.

[0076] FIG. 4 shows the particle size distribution of all particles and their change by way of the different post-treatments.

[0077] It can be recognised that for the solution-based LNMO particles (AT_L) which are manufactured in the pulsating hot gas flow, no changes in the particle size distribution occur due to a post-treatment. The suspension-based LNMO particles (AT_S) which are manufactured in the pulsating hot gas flow exhibit different particle size distributions. Two particles sizes, specifically around 1 μm and around 30 μm manifest themselves without post-treatment (AT_S_0) and with a single-stage post-treatment (AT_S_5). By way of the mechanical post-treatment in the form of the grinding (AT_S_5g), the particle size distribution is unified at a particle size of around 2.5 μm. Hence the solution-based LNMO particles which are manufactured in the pulsating hot gas flow have a narrower particle size distribution than the suspension-based LNMO particles which are manufactured in the pulsating hot gas flow. A change of the particle size distribution is shown in the same manner for the spray-dried LNMO particles (ST_L). This being from a wide distribution with two peaks for the untreated particles of the LNMO particles (ST_L_0) to a much narrower distribution with a peak on account of the 2-stage post-treatment (ST_L_5g). The post-treatment breaks up aggregates and thus leads to a more homogeneous particle size distribution.

[0078] The d90 diameter and the specific surface area which was determined via the BET method are shown in FIG. 5. The results are represented in three columns for the different methods (AT_L, AT_S, ST_L) and per column for the three different types of post-treatments (0, 5, 5g).

[0079] The particle size distribution (d90) is the same for solution-based particles of the LNMO particles (AT_L) which are manufactured in a pulsating hot gas flow. A change can be recognised in the specific surface area. The specific surface area of the LNMO particles is greatly reduced from 20 m.sup.2/g to 4 m.sup.2/g by way of the single-stage post-treatment in the form of an additional calcination. No significant change in comparison to the single-stage post-treatment is achieved by way of the 2-stage post-treatment.

[0080] The d90 diameter lies at round 40 μm for LNMO particles which were subjected to no or a single-stage post-treatment. In contrast to this, the d90 diameter for LNMO particles which were subjected a 2-stage post-treatment lies at around 5 μm. For the suspension-based particles (AT_S) which are manufactured in a pulsating hot gas flow, the specific surface area without an additional treatment (AT_S_0) at 7 m.sup.2/g in comparison is lower than in the case of solution-based particles (AT_L_0). By way of a single-stage post-treatment (AT_S_5), the specific surface area is reduced to 1 m.sup.2/g and by way of a 2-stage post-treatment to 2 m.sup.2/g.

[0081] The spray dried particles as a whole have specific surface area values around 1-3 m.sup.2/g. The particle size distribution (d90) reduces from no post-treatment (ST_L_0) via a single-stage post-treatment (ST_L_5) to a 2-stage post-treatment (ST_L_5g) from 38 to 6.

[0082] The additional grinding (*_5g) has an influence on the particle size distribution, wherein the specific surface areas remain the same. Considered as a whole, a post-treatment leads to a reduction of the particle size distribution and of the specific surface area.

[0083] The measurement of the particle size distribution (laser diffraction) and of the specific surface area (BET) do not provide any information on the individual surface of the primary particles; in contrast the results shows the aggregation and the agglomeration of the particles. The particles (AT_L, AT_S) which are produced by way of the method of a pulsating hot gas flow exhibit a high specific surface area without post-treatment. A recognisable crystal growth does not take place, which is why the specific surface area continues to remain large, since the structure for the most part is influenced by way of the fine distribution during the spraying procedure. The crystals begin to shape and the specific surface area reduces on account of the thermal post-treatment. No significant change is perceived for the spray-dried particles (ST_L) since the crystals have already been shaped. In contrast, the change in the particle size distribution is very different. It should be noted that the LNMO particles have been treated differently for the analysis methods. The powder was dissolved in a solution for the particle size distribution. Agglomerates were broken up and for this reason differences in the BET method could be recognised where agglomerates continue to exist. The particle size distribution did not change for the solution-based particles (AT_L) which are produced in the pulsating hot gas flow, wherein the agglomerates were destroyed by way of the post-treatment. The suspension-based particles (AT_S_5) which are manufactured in a pulsating hot gas flow and having a single-stage post-treatment exhibit no significant differences in the particle size in comparison to the untreated particles (ST_L_0). The particle size could be reduced exclusively by way of a 2-stage post-treatment, in particular by way of the grinding step which breaks up the aggregates. With regard to the specific surface area, the characteristics of the suspension-based particles (AT_S) are similar to those of the solution-based particles (AT_L), which leads to a similar conclusion with regard to the formation of the crystals. For the spray-dried particles (ST_L), a temperature post-treatment leads to smaller crystals. This can be explained by the fact that larger aggregates form smaller crystals. This change is additionally shown in the FESEM pictures in FIG. 3.

[0084] The electrochemical characterisation and the galvanostatic cycle characteristics are shown in the FIGS. 6 to 8 and Table 2. The battery power and the cycle stability over different C-rates were characterised with the galvanostatic method. A discharge current with a rate of “1 C” herein corresponds to the current which is necessary in order to discharge the battery cell in one hour. Accordingly for example, a discharge current of “2 C” corresponds to a current which is necessary in order to discharge the battery cell in half an hour and a discharge current of “½C” or “0.5 C” corresponds to a current which is necessary to discharge a battery cell in two hours.

[0085] The rate capability is shown for all LNMO particles in FIG. 6. The particles exhibit a significant difference in the specific capacity at the measured C-rates. By way of a single-stage post-treatment in the form of an additional calcination, the values are improved significantly, as is shown in FIG. 6. The grinding step, the 2-stage post-treatment also has a positive influence on the specific capacity. However, one can also recognise that the increase of the specific capacity is different for different particles. It is particularly the suspension-based particles (AT_S_0) which are manufactured in the pulsating hot gas flow which at 41 mAh/g have a very low capacity without an additional post-treatment, whereas the solution-based particles of LNMO (AT_L_0) which are manufactured in the pulsating hot gas flow and the spray dried particles (ST_L_0) already have 100 mAh/g and 81 mAh/g respectively.

[0086] After a thermal treatment for the suspension-based LNMO particles (AT_S_5) which are manufactured in a pulsating hot gas flow, the specific capacity can be more than doubled from 27% to 60%, which is likewise the case for solution-based particles of the LNMO particles (AT_L_5) which are manufactured in a pulsating hot gas flow and the spray-dried particles (ST_L_5).

[0087] The additional grinding step which is contained in the 2-stage post-treatment has no influence on the spray-dried particles (ST_L_5g). In contrast to this, this however increases the specific capacity for the particles (AT_L_5g, AT_S_5g) which are manufactured in the pulsating hot gas flow, as is further shown in Table 2. The specific capacity for different C-rates is shown in FIG. 6 and is summarised in Table 2. In general, the LNMO particles which were subjected to a 2-stage post-treatment have a lower capacity loss at increasing C-rates. At 0.1 C, a low electrical charging, the discharge time of the battery permits a complete intercalating of lithium ions in the anode and leads to a maximal utilisation of the battery capacity during the charging/discharging cycles. In contrast to this, at 5 C, a high electrical charging, the rapid discharging times do provide the battery with sufficient time in order to render the maximal capacity usable. This phenomena is exhibited by all particles, but differs significantly with regard to the material characteristics and the manufacturing method of LNMO particles. The capacity correlates to the particle size distribution. The electrolyte is in the position of achieving high quantities of crystalline active material.

TABLE-US-00002 capacity spec. increase ageing capacity with post- 0.1 C 0.1 C 0.1 C after Mn3 +− at 0.1 C treatment zu 1 C zu 2 C zu 5 C 20 cycles 0.1 C share [mAh/g] [%] [%] [%] [%] [%] Recyling [%] AT_L_0 100 — 12 35 89 5 11 50 AT_L_5 127 27 3 7 46 3 8 9 AT_L_5g 136  7 3 5 13 3 7 8 192 41 — 29 66 98 −30 5 86 AT_12S_5 120 192  1 2 24 1 2 47 AT_S_5g 134 12 0 1 10 1 1 23 ST_L_0 81 — 5 9 26 1 −4 34 ST_L_5 130 60 0 0 5 1 1 12 ST_L_5g 130  0 0 0 3 1 1 8

[0088] The specific surface area in m.sup.2/g and the particle size distribution (d90) is correlated with regard to the measured specific discharge capacity in mAh/g at a rate of 0.1 C in FIG. 7. In order to describe the correlation, the change of the specific surface area and the particle size are represented in a combined manner and as a black line. The change of the specific discharge capacity at 0.1 C is specified by the red line. A reduction of the specific surface area and particle size leads to an increase in the specific discharge capacity. Both material characteristics influence the electrochemical characteristics, as shown, since the specific discharge capacity is very small in the case of a very large specific surface area and particle size. This is particularly recognisable with the particles without post-treatment (*_0). The largest specific discharge capacity is achieved by the particles by an additional grinding step of the 2-stage post-treatment (*_5g), which exhibit values of greater than 130 mAh/g. A small particle size and specific surface area was achieved for these LNMO particles.

[0089] Given an increase in the C-rate from 0.1 C to 1 C, the suspension-based particles (AT_S) and the spray-dried particles (ST_L) exhibited no significant reduction of the capacity with additional treatment steps. This likewise applies to an increase to 2 C, wherein here it is only the suspension-based particles (AT_S) which exhibit a small drop of 1-2%. At a C-rate of 5 C, the capacity drops significantly for all particles except for the spray-dried particles (ST_L_5, ST_L_5g) which exhibit the smallest reduction by only 3% and 5% respectively. Table 2 likewise shows the ageing and the specific capacity after the complete program. In this case, the ground particles (ST_L_5g) have only lost 1% of their initial capacity at the end of the trial series (after 38 cycles), which is little compared to the other particles. A large reduction of the capacity was recognized for particles of the LNMO particles which were subjected to no additional calcination (AT_L_0, AT_S_0, ST_L_0). The particles without additional post-treatment are greatly influenced by the increase of the C-rate. The material characteristics of these particles play an important role in the reduction of the specific capacity given an increase of the C-rate. The material with larger crystals exhibits less reduction of the specific discharge capacity than the material with small crystals. The size differences can be recognised in FIG. 3.

[0090] This leads to the fact that the LNMO particles with a larger surface and/or with a smaller crystal size assume a greater interaction with the electrolytes during the charging/discharging cycles, which subsequently results in a deterioration of the capacity. The LNMO particles which have a uniform particle size distribution according to FIG. 4 exhibit a lower reduction of the capacity than the LNMO particles with a non-uniform distribution. A more uniform particle size distribution correlates to a reduction of very large crystals which are not suitable for intercalating and de-intercalating lithium. Additionally, in the literature it is noted that an octahedral crystal shape has improved electrochemical characteristics (FIG. 3) in comparison to other crystal shapes. Apart from the material characteristics, the reduction of the specific capacity is also influenced by the different manufacturing methods. If the different post-treatments in FIG. 6 are considered, a reduction of the capacity for spray-dried particles (ST_L) is smaller than for particles (AT_*) which are manufactured in the pulsating hot gas flow. As already mentioned, due to an extended calcination time with spray pyrolysis, an adequate time is made available for forming the crystal planes during the primary calcination (STT_L), this corresponding to the spray drying. In contrast, with the method with a pulsating hot gas flow, the sojourn time in the reactor is very much smaller—less than 10 seconds—which is why a formation of the crystal planes by way of primary calcination does not take place, which subsequently leads to an increased reduction of the specific discharge capacity.

[0091] The number of Mn.sup.3+ ions can be estimated on the basis of the plateau in the voltage region from 3.5 V to 4.5 V of the charging/discharging voltage curves in FIG. 8. It is assumed that given a voltage of 4.5 V, all Mn.sup.3+ ions which are located in the material are oxidised into Mn.sup.4+ ions. The percentage of Mn.sup.3+ ions of all particles was computed from FIG. 8 and summarised in Table 2. According to FIG. 8, the LNMO particles (AT_L_0, AT_S_0) which are manufactured in a pulsating hot gas flow comprise very large quantities of 50% to 86% of Mn.sup.3+ ions which cover more than half the voltage curve in the voltage region between 3.5 and 4.5 V, wherein the voltage plateau at 4.7 V is much more dominant in all other LNMO particles. The LNMO particles which comprise a large number of Mn.sup.3+ ions (AT_L_0, AT_S_0 and ST_L_0) consequently exhibit a poor energy density. Furthermore, the post-treatments reduce the share of Mn.sup.3+ ions. The main reason for the grinding step before the additional calcination is to increase the oxygen interaction during the additional calcination. This leads to an improved crystal growth and to the reduction of the quantity of Mn.sup.3+ ions. The suspension-based particles (AT_S_0) at 86% have a large quantity of Mn.sup.3+ ions, wherein the other untreated particles (AT_L_0, ST_L_0) comprise a quantity of Mn.sup.3+ ions of 50% to 34%. Additionally, during the additional calcination, the percentage of Mn.sup.3+ for suspension-based particles (AT_S_5) was significantly reduced to 46%, for the solution-based particles (AT_L_5) manufactured in a pulsating hot gas flow to 8.5% and for spray-dried particles (AT_L_5) to 11.7%. The additional grinding has a low influence on the reduction of the Mn.sup.3+ ions in the case of the solution-based particles (AT_L_5g, ST_L_5g), wherein a reduction of more than 50% takes place for suspension-based particles (AT_S_5g) (Table 2). The suspension-based LMNO particles (AT_S) which are manufactured in the pulsating hot gas flow have a larger quantity of Mn.sup.3+ ions than other particles. The cause of this is the contact surface between the individual precursor particles with oxygen particles being smaller during the calcination in comparison to the solution-based LNMO particles. This leads to a higher Mn.sup.3+ ion content for suspension-based LNMO particles than for solution-based LNMO particles.

[0092] The aforementioned correlations between the material characteristics and electrochemical characteristics shows that an examination of the phenomenon of the electrolyte decomposition is necessary given high voltages (4.7 V vs. Li/Li.sup.+). The electrolyte decomposition is due to the undesirable side reactions which take place in the cell, which leads to a larger quantity of reduced substances on both electrodes. This effects a block for the lithium ion transport. The undesirable reactions occur in all particles, but differ enormously in regard to the material morphology (crystal presence, crystal size and crystal structure).

[0093] The cumulative charge and discharge end point capacities of all particles which are measured during the galvanostatic measurements are shown in FIG. 9. During the changing loading of the cells, the charging and discharging curves do not exactly correspond due to the parasitic currents, and the curves consequently shift from one cycle to the next. Generally, it can be said that the more the curves shift to the right, the more parasitic currents have flowed. The shifting between the cycles is understood as a capacity shift. The cumulative capacity represents the specific capacity of the respective cycle just as the shifting of each preceding cycle. The value of the cumulative end point capacities increases steeply up to the thirteenth to fifteenth cycle, which belongs to a C-rate of 5 C. The cumulative end point capacity subsequently increases less steeply from the sixteenth cycle to the thirty-eighth cycle.

[0094] Although these particles were charged and discharged at different C-rates, the quantity of foreign reactions which occur in the cells continues to be reliably differentiated by way of considering the magnitude of the end point capacities (y-axis) at the end of the cycles. The ratio of the magnitude of the capacity shift (gradient of the curve) is very much larger at the initial cycles (1 to 3) than in the middle and at the end of the cycles, which is most probably due to the formation of the SEI (solid electrolyte interphase) which occurs to a high degree in the initial cycles, by which means the growth rate is then significantly reduced in the subsequent cycles. The size of the foreign reactions is also highly dependent on the contact duration at high voltages of the electrode with respect to the electrolyte. The discussed values after 38 cycles were used for this.

[0095] In FIG. 9, the untreated suspension-based particles (AT_S_0) exhibit a very high cumulative discharge end point capacity without an additional treatment of 370 mAh/g whereas the solution-based particles (AT_L_0, ST_L_0) have a value of 211 mAh/g and 86 mAH/g respectively. After an additional calcination, the cumulative discharge end point capacities reduce significantly to 80 mAh/g for suspension-based particles (AT_S_5), to 104 mAh/g for solution-based particles (AT_L_5) which are manufactured in a pulsating hot gas flow and 68 mAh/g for spray-dried solution-based particles (ST_L_5).

[0096] The additional grinding has essentially no influence on the cumulative discharge end point capacity for the particles (AT_L_5g, AT_S_5g) which are manufactured in the pulsating hot gas flow, but the cumulative discharge end point capacity increases slightly for the spray-dried particles (ST_L_5g), as is shown in FIG. 9. The same also applies to the cumulative charging end point capacity.

[0097] From FIG. 9, it is evident that the cumulative discharge end point capacities behave linearly to the specific surface area of the particles. This lies in the high electrolyte and electrode contact surface when the material has a larger surface. The larger contact surface leads to a larger electrolyte decomposition and subsequently to a greater capacity shift.

[0098] FIG. 10 shows the values of the cumulative discharge end point capacities in the magnitude of 10 [×10]. Apart from the specific surface area, the presence of crystals has a great influence on the capacity shift. The spray-dried particles (ST_L) have a reduced capacity shift compared to the other particles (AT_S, AT_L). The reason for this is the presence of crystal planes after the primary calcination. No crystal planes are present after the primary calcination in the case of particles which are manufactured in the pulsating hot gas flow. According to these correlations, it is recommended to keep the crystal size at 1 μm in order to maintain an optimal electrochemical performance and reduced electrolyte decomposition. From this, one can conclude that the electrolyte decomposition is greatly influenced by the specific surface area of the material (FIG. 10). A uniform crystal size distribution and suitable crystal sizes are necessary in order to reduce a deteriorating capacity and the kinetic hindrance.

[0099] EIS provides information on the kinetic characteristics for the lithium ion transport in the electrodes. The capability of the lithium ion transport controls the kinetic characteristics and the electrode reaction, wherein this factor is greatly influenced by the material morphology. The sequence of the transport also encompasses the lithium ion transport process, the electron transport process and the charge transfer process. On account of the differences in their time constants between these processes, EIS is a suitable technology in order to examine these reactions and can permit such phenomena to be separated. Whilst using EIS and equivalent circuit analyses therefore, one can examine the kinetic parameters which are associated with the lithium ion integration and removal in intercalation materials, such as the surface film resistance, the charge transfer resistance and the electrolyte resistance. The varying impedance was applied at the difference state of charge of the galvanostatically cycled button cells. It is evident from FIG. 11 that the width of the semicircles varies for the materials with a different morphology. All particles at an SOC of 90% differ marginally from the other states of charge. At 90% SOC, the particles ST_L_0 have a very low impedance value without an additional treatment, in comparison to the particles (AT*) which are manufactured in the pulsating hot gas flow. This is down to the presence of crystal planes after the primary calcination, whereas the particles which are manufactured in the pulsating hot gas flow, without additional treatment (AT_L_0, AT_S_0) have no crystals after the primary calcination (FIG. 3). As is shown in FIG. 11, the impedance values are significantly reduced after the post-treatment. According to the interpretation of the material characteristics and electrochemical characteristics, the impendence values are greatly dependent on the crystal size and the crystal size distribution. The particles which have a homogenous crystal distribution and larger crystals (˜1 μm) exhibit reduced impedance values compared to other particles.

[0100] In order to further analyse this information with regard to impedances and to understand the correlation between material morphology and the kinetic characteristics, an equivalent circuit analysis was carried out. The information on the electrolyte resistance, the surface film resistance and the charge transfer resistance can be obtained by way of modelling with an equivalent circuit diagram according to FIG. 12.

[0101] FIGS. 13 to 15 show the results of the adaptation with an equivalent circuit diagram. The electrolyte resonance, the surface film resistance and the charge transfer resistance are the parameters which are considered for the correlation of the material characterises with the kinetic characteristics.

[0102] FIG. 13 shows the surface film resistance at different SOCs. The surface film resistance of the material represents the stability of the cathode material surface with respect to the reactive electrolyte. The surface film resistance values differ in accordance with the surface film thickness, and from this interpretation it is possible to estimate the size magnitude of the electrolyte breakdown of different materials. FIG. 13 shows that the surface film resistance (Rsf) at 10% SOC is very much larger than with other SOCs. It is speculated that the passivation film is predominantly formed at the initial SOC and remains almost stable during the SOC increase. FIG. 13 shows that given a SOC of 10%, the spray-dried particles (ST_L_0) have a very low surface film resistance at 14Ω without additional treatment, whereas that of the particles which are manufactured in the pulsating hot gas flow without post-treatment (AT_S_0, AT_L_0) is 73Ω and 59Ω respectively. Furthermore, during the additional calcination, the surface film resistance values for particles which are manufactured in the pulsating hot gas flow are significantly reduced at 23Ω (AT_S_5) and 29Ω (AT_L5) and the spray-dried particles (ST_L_5)/ST_L_5) have a similar value of 17Ω. The grinding step has further reduced the surface resistance, as is shown in FIG. 13. The same interpretation can also apply to other SOCs. The correlation to the specific surface area is illustrated in FIG. 16. Here, it is shown that the reduction of the surface leads to a reduction of the surface film resistance. Generally, from the different electrochemical methods according to Table 2, one can conclude that the magnitude of the foreign reactions in the cell is directly proportional to the specific surface area.

[0103] FIG. 14 shows the charge transfer resistance at different SOCs. The charge transfer at an electrode/electrolyte boundary surface is an essential procedure of the charging/discharging reaction of lithium ion batteries. This phenomenon determines the speed of the reactions in the electrode. Generally, the charge transfer values are greatly influenced by the SOC and this phenomenon is varied according to the different particles. FIG. 14 at 90% SOC shows that the untreated spray-dried particles without additional treatment (ST_L_0) have a very low charge transfer resistance of 3Ω, whereas that of the particles (AT_S_0, AT_L_0) which are manufactured in the pulsating hot gas flow in each case is 17Ω. The charge transfer resistance value for the particles (AT_S_5, AT_L_5) which are manufactured in the pulsating hot gas flow has reduced significantly to 4Ω during the additional calcination, and has increased to 9Ω for the spray-dried particles (ST_L_5). The grinding step has further reduced the charge transfer resistance for all particles, as is shown in FIG. 14. The same trend can be applied to other SOCs which are represented in FIG. 14. The charge transfer resistance is directly influenced by the particle size and pore distribution within the electrode layer.

[0104] The capability of the charge transfer resistance is controlled by the contact of particle to particle and by way of the continuous crystal network within the electrode layer. And this continuous crystal network provides an improved electrode and electrolyte boundary surface within the porous electrode. As expected, the material without crystals (AT_L, AT_S and ST_L) has a higher charge transfer resistance than the other materials.

[0105] FIG. 15 shows the electrolyte resistance at different SOCs. According to the electrolyte resistance (Re), the conductivity (σ) can be estimated by the thickness of the electrolyte film (electrolyte-impregnated separator thickness 0.052 cm) and by the electrode contact surface to the electrolyte (0.013 cm.sup.3). Generally, the electrolyte resistance of all particles lies between 3Ω to 7Ω. According to the results, the electrolyte resistance is predominantly dependent on the particle size of the material and the distribution. Disregarding the material morphology, the difference of the values could be down to the handling of the button cells on construction. When the button cell is put together, there exists the possibility of losing a very small quantity of electrolyte in the inner housing of the button cell.