Glass-coated cathode powders for rechargeable batteries

Abstract

The invention provides a cathode active material for use in a rechargeable battery, comprising a coated lithium nickel oxide powder or a coated lithium nickel manganese oxide powder, the powder being composed of primary particles provided with a glassy lithium silicate surface coating. A method for preparing the cathode active material comprises the steps of: providing a lithium transition metal based oxide powder, providing an alkali mineral compound comprising a Li.sub.2xSiO.sub.30.5x compound, wherein 0<x<2, mixing the lithium transition metal based oxide powder and the alkali mineral compound to form a powder-mineral compound mixture, and heat treating the mixture at a temperature T whereby lithium is extracted from the surface of the metal based oxide powder to react with the alkali mineral compound, and a glassy surface coating is formed comprising a Li.sub.2xSiO.sub.30.5x compound, wherein x<x<2.

Claims

1. A cathode active material for use in a rechargeable battery, comprising a coated lithium nickel oxide powder or a coated lithium nickel manganese oxide powder, the powder comprising primary particles having a glassy surface coating, wherein the coating comprises a lithium silicate compound, wherein the lithium silicate compound has lithium accepting properties, and wherein the coating comprises a compound having a chemical composition expressed by Li.sub.2xSiO.sub.30.5x, wherein 0<x1.6.

2. The cathode active material of claim 1, wherein the glassy surface coating further comprises either one or both of a phosphate and borate compound, said compound having lithium accepting properties.

3. The cathode active material of claim 1, wherein the glassy surface coating comprises lithium.

4. The cathode active material of claim 2, wherein the glassy surface coating comprises either one or both of a Li.sub.32yPO.sub.4y and a Li.sub.32zBO.sub.3z compound, wherein 0<y<1.5 and 0<z<1.5.

5. The cathode active material of claim 1, wherein the glassy coating compound has a composition gradient, wherein the value of x at the surface of the primary particles is lower than the value of x at the outer surface of the glassy coating.

6. The cathode active material of claim 1, wherein the cathode active material comprises between 0.07 and 1 wt % of Si.

7. The cathode active material of claim 2, wherein the cathode active material comprises between 0.1 and 2 wt % P.

8. The cathode active material of claim 2, wherein the cathode active material comprises between 0.03 and 0.5 wt % B.

9. The cathode active material of claim 1, wherein the glassy surface coating consists of nano-composites of Li.sub.2Si.sub.5O.sub.11 and Li.sub.2SiO.sub.3 particles.

10. The cathode active material of claim 1, wherein the cathode active material comprises between 0.05 and 0.5 mol % glassy surface coating.

11. The cathode active material of claim 1, wherein the primary particles are either one of Li.sub.aNi.sub.xCO.sub.yN.sub.zO.sub.2eA.sub.f, with 0.9<a<1.1, 0.5<x0.9, 0<y<0.4, 0<z<0.35, e<0.02, 0<f<0.05 and 0.9<(x+y+z+f)<1.1; N consisting of one or more elements selected from the group consisting of Al, Mg, and Ti; and A consisting of one or both of S and C; and Li.sub.1+aM.sub.1-aO.sub.2bM.sub.kS.sub.m with 0.03<a<0.06, b<0.02, wherein at least 95% of M=Ni.sub.aMn.sub.bCo.sub.c, with a>0, b>0, c>0 and a+b+c=1; and a/b>1; wherein M consists of one or more elements selected from the group consisting of Ca, Sr, Y, La, Ce and Zr, with 0<k<0.1 in wt %; and wherein 0<m0.6, m being expressed in mol %.

12. The cathode active material of claim 11, wherein the primary particles are Li.sub.1+aM.sub.1-aO.sub.2bM.sub.kS.sub.m with M=Ni.sub.aMn.sub.bCo.sub.c, and wherein 1.5<a/b<3, and 0.1c<0.35.

13. The cathode active material according to claim 12, wherein 0.5<a<0.7.

14. A method for preparing the cathode active material of claim 1, comprising: providing a lithium transition metal based oxide powder, providing an alkali mineral compound comprising a Li.sub.2xSiO.sub.30.5x compound, wherein 0<x<2, mixing the lithium transition metal based oxide powder and the alkali mineral compound to form a powder-mineral compound mixture, and heat treating the mixture at a temperature T between 300 and 500 C., whereby a glassy surface coating is formed comprising a Li.sub.2xSiO.sub.30.5x compound, wherein x<x<2.

15. The method according to claim 14, wherein the heat treatment is performed in an oxygen comprising atmosphere.

16. The method according to claim 14, wherein the alkali mineral compound is provided as a dry nanometric powder; and during the heat treatment of the mixture, the powder is sintered and adheres to a surface of the transition metal based oxide powder in the form of a glassy coating.

17. The method according to claim 14, wherein the alkali mineral compound is provided as an aqueous solution of the alkali mineral compound; and during the heat treatment of the mixture, water from the aqueous solution evaporates and the compound dries to form a glassy coating on a surface of the metal based oxide powder.

18. The method according to claim 14, wherein the lithium transition metal based oxide powder consists of either one of Li.sub.aNixCo.sub.yNzO.sub.2eA.sub.f, with 0.9<a<1.1, 0.5<x<0.9, 0<y<0.4, 0<z<0.35, e<0.02, 0<f<0.05 and 0.9<(x+y+z+f)<1.1; N consisting of one or more elements selected from the group consisting of Al, Mg, and Ti; and A consisting of one or both of S and C; and Li.sub.1+aM.sub.1-aO.sub.2bM.sub.kS.sub.m with 0.03<a<0.06, b<0.02, wherein at least 95% of M=Ni.sub.aMn.sub.bCo.sub.c, with a>0, b>0, c>0 and a+b+c=1; and a/b>1; wherein M consists of one or more elements selected from the group consisting of Ca, Sr, Y, La, Ce and Zr, with 0<k<0.1 in wt %; and wherein 0m<0.6, m being expressed in mol %.

19. The method according to claim 18, wherein the lithium transition metal based oxide powder consists of Li.sub.1+aM.sub.1-aO.sub.2bM.sub.kS.sub.m with M=Ni.sub.aMn.sub.bCo.sub.c, and wherein 1.5<a/b<3, and 0.1c<0.35.

20. The method according to claim 19, wherein 0.5a<0.7.

21. The method according to claim 14, wherein the alkali mineral compound consists of Li.sub.2Si.sub.5O.sub.11 or Li.sub.2Si.sub.2O.sub.5.

22. The method according to claim 14, wherein the heat treatment of the mixture is performed at a temperature T between 350 and 450 C. for at least one hour.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 shows a SEM micrograph of uncoated cathode precursor.

(2) FIG. 2 shows the properties (base content in mol/g, capacity in mAh/g, cycle stability in % per 100 cycles) of coated 622 samples as function of treatment temperature ( C.).

(3) FIG. 3.1 shows a high resolution SEM of the Li-silicate coated EX sample, heat-treated at 200 C.

(4) FIG. 3.2 shows a high resolution SEM of the Li-silicate coated EX sample, heat-treated at 400 C.

(5) FIG. 3.3 shows a high resolution SEM of the Li-silicate coated EX sample, heat-treated at 600 C.

(6) FIG. 4 shows an X-ray diffraction pattern of Li.sub.2Si.sub.5O.sub.11 liquid glass after drying at 200, 400, 600 C., respectively.

(7) FIG. 5 shows an X-ray diffraction pattern of a mixture of dried Li.sub.2Si.sub.5O.sub.11 glass and Li.sub.2CO.sub.3 before (top graph) and after heat treatment at 450 C. for 72 hours (bottom graph).

DETAILED DESCRIPTION OF THE INVENTION

(8) The powderous cathode material of the current invention is a coated material. The surface of the powder particles may be fully covered by a thin coated film. The coating layer is a secondary coating, it is different from the initially applied primary or pristine coating. Aspects of the primary and secondary coating are described in the following:

(9) A dense surface coverage: a successful coating my require a full surface coverage by the coating film. Full surface coverage may be achieved by either of the following methods:

(10) 1) Dry (powder) coating followed by a heat treatment which involves the melting of the dry powder, to provide a liquid with excellent wetting properties, whereafter the surface will become homogeneously coated by a thin liquid film. After cooling to a temperature below the melting point a solid, dense coating film is obtained.
2) Dry nano powder coating followed by a heat treatment which involves a soft sintering of the dry nano powder, whereby the powder adheres to the surface and forms a continuous protective film.
3) Wet coating: using a liquid having excellent wetting properties, the surface of the powder becomes homogeneously coated by a thin liquid film. The film is then solidified by drying-evaporating or cooling, allowing to obtain a dense coating film.

(11) In an embodiment of the invention, the primary coating film of the materials is obtained by a wet coating and drying-evaporating approach. However only few liquids allow to obtain a dense surface coating after drying-evaporating the liquid. We discovered that the formation of a crystalline solid after drying is highly undesired, since during crystal formation large parts of the surface can become uncoated. In this embodiment a solution is used which after drying-evaporating results in a glassy coating. We observed that by using a glass forming liquid excellent surface coverage can be obtained. Glass forming liquids are for example solutions of poly-phosphate, -borates or -silicates. Polysilicates of alkali minerals are called water glass, they are compounds that are transparent solids, having high melting points (above 800 C.) and being water soluble. A known example of water glass is a solution of disodium pentasilicate Na.sub.2Si.sub.5O.sub.11 in water. It is not clear if Na is tolerated in Li batteries so one embodiment focuses on the Li substitute lithium polysilicate Li.sub.2Si.sub.5O.sub.11. Other examples of liquid glass are polyphosphates, such as LiPO.sub.3 solutions in water, or polyborates like LiBO.sub.2 solutions.

(12) To summarize: this embodiment provides the use of liquid glass for the pristine wet coating step. We refer to liquid glass and water glass as a solution which after evaporation and drying forms a glass and not a crystal. A glass is an amorphous solid, which is capable of passing continuously into the viscous liquid state. The use of liquid or water glass allows to obtain a dense, thin inorganic coating layer.

(13) A low base content obtained by a lithium acceptor: The current invention utilizes a unique feature: the pristine glass coating may be a strong lithium acceptor. For example, in the embodiment described before, after a liquid coating of the cathode powder and drying-evaporating to form a solid glassy coating layer, the base content is similar or even higher than for the original cathode powder. At relatively low temperature it has a strong desire to react with lithium. The ability to react already at low temperature with surface lithium is surely related to the good surface wetting ability of the glassy coating. The soluble surface base of the cathode powder may contain lithium (but it is not an impurity in the sense of a secondary phase, it is merely a surface property). Thus the surface base is a lithium donator. As an example: after immersing the powder into water Li containing surface compounds dissolve, as a result Li.sub.2CO.sub.3 and LiOH dissolve and the pH increases. The dissolved lithium compounds originate from the soluble base. The lithium on the surface is thermodynamically less bound than the lithium in the bulk. Therefore it is possible to remove lithium from the surface but not from the bulk by using suitable techniques. A suitable technique is the surface delithiation by a glassy coating layer which has Li accepting properties. Example reactions are:
LiPO.sub.3+2LiOH.fwdarw.Li.sub.3PO.sub.4+H.sub.2O,
Li.sub.2Si.sub.5O.sub.11+8LiOH.fwdarw.5Li.sub.2SiO.sub.3+4H.sub.2O, or
LiBO.sub.2+2LiOH.fwdarw.Li.sub.3BO.sub.3+H.sub.2O.

(14) In an embodiment of the invention, a controlled heat treatment of the primary glassy coated powder allows to decompose the surface base without attacking the bulk. A typical treatment temperature is 300-500 C. At lower temperatures the surface base is not sufficiently decomposed. At higher temperature the delithiation reactions continues beyond the decomposition of surface base, and as a result the bulk phase of the cathode powder is attacked. During further reaction the glass would decompose, lithium would be extracted from the bulk until a high Li stoichiometric crystalline product (Li.sub.3PO.sub.4, Li.sub.3BO.sub.3 or Li.sub.2SiO.sub.3) is formed, and the cathode material also re-creates the surface base. Contrary to these high Li stoichiometric crystalline products, the silicate, phosphate and borate compounds having lithium accepting properties of the present invention are low lithium stoichiometric silicate, phosphate or borate compounds.

(15) We speculate that the narrow temperature range is related to the reactivity of surface and bulk oxygen. Lithium is very mobile already at room temperature (otherwise the cathode material would not be an intercalation material). However, de-intercalated cathode materials are thermodynamically unstable. A chemical deintercalation is only possible with very strong oxidizers, and deintercalated compounds collapse at sufficient high temperature (about 400 C.) and release oxygen.

(16) For example:

(17) Li.sub.1+xM.sub.1xO.sub.2.fwdarw.de-intercalation of 2x Li.fwdarw.Li.sub.1xM.sub.1xO.sub.2.fwdarw.(1x)LiMO.sub.2+x O.sub.2. So, the reaction of bulk lithium is impossible as long as the bulk oxygen is immobile, whereas under similar conditions surface oxygen and surface base might already be reactive. This explains the relative narrow T range where a reduction of only surface base by Li accepting glassy coating happens.

(18) The coating film of the final product originates from a glass, but it is not the pristine applied coating. The coating film is the result of the reaction of the glass with lithium. The lithium is supplied by the Li containing soluble surface base. Thus by the reaction of the glass with lithium the surface base is decomposed and the amount of soluble base is dramatically decreased. Therefore the cathode materials of the present invention have excellent high temperature storage properties when charged batteries are exposed to heat.

(19) We believe that the secondary coating is a double shell coating where the outside of the coating layer still has the pristine glass composition but the inside shell has a higher Li stoichiometry, eventually being a nano-composite of pristine glass (for example Li.sub.2Si.sub.5O.sub.11) and small amounts of the lithiated phase (for example Li.sub.2SiO.sub.3).

(20) To summarize: an important aspect of the current invention is that the pristine glassy surface layer has Li accepting properties, and by applying a controlled temperature treatment within a narrow temperature range the glassy coating layer partially reacts with the surface base, thereby forming a secondary coating and consuming the surface base.

(21) Soluble surface base: for the evaluation of slurry stability and stability during high temperature storage of final cellsas discussed beforethe soluble base is an important factor. In the following we will explain soluble base and the mechanism to decrease the soluble base using experimental examples. If the cathode powder is immersed in water the surface compound dissolves and causes an increase of pH, thus we call it soluble base. Lithium located near to the surface is thermodynamically less stable than lithium in the bulk. It can dissolve (by ion exchange) or it can react with molecules within the atmosphere. Contrary to this, the lithium in the bulk is thermodynamically more stable since it cannot be dissolved, and hence is less reactive.

(22) As the lithium near to the surface is reactive, in the most simple case it will bind atmospheric oxygen to the surface, forming an oxygen-lithium surface compound. If the atmosphere contains moisture then hydrogen-oxygen-lithium surface compounds can form. If the cathodes containing these surface compounds are immersed in water the surface compound will dissolve. In the case of oxygen-lithium and in the case of hydrogen-oxygen-lithium surface compounds the dissolved compound is lithium hydroxide. In a more complex case the atmosphere contains carbon, for example in the form of carbon dioxide or organic radicals. Then the surface compoundsbesides of oxygen and lithium, eventually hydrogenalso contains carbon. If the cathodes with a carbon containing surface compound are immersed the compound dissolves and lithium carbonate is formed. Additionally Li near to the surface can go dissolve by an ion exchange reaction Li+.fwdarw.H+. All these reactions form dissolved bases in the form of LiOH or Li.sub.2CO.sub.3. So the soluble base is not an impurity but it is rather a surface property, it is the ability of the cathode to perform the above mentioned reactions by the presence of reactive lithium surface compounds.

(23) The amount and composition of the base can be qualitatively (hydroxide versus carbonate) and quantitatively (mol/g cathode) determined by pH titration. In the pH titration the cathode is immersed in water, the soluble base dissolves, and after filtering, the amount and type of dissolved base is obtained by monitoring the pH profile. This mechanism is explained in co-pending application EP11000945.3. If all soluble base is dissolved (that means all reactive surface lithium compounds have reacted with water) then typically the creation of more surface compounds is stopped (or slowed down) because lithium in the bulk is thermodynamically more stable than Li on the surface. Practically the lithium which goes to dissolution is a surface property and not an impurity. If the soluble surface compounds are removed and the sample is dried and reheated then the soluble base (meaning reactive lithium surface compounds) is restored. The restoration easily happens in the case of LNMC as will be explained next.

(24) Lithium transition metal oxides which contain manganese and nickel have a Lithium non stoichiometric range. As an example, Li.sub.1+xM.sub.1xO.sub.2 with M=Ni.sub.0.5Mn.sub.0.3Co.sub.0.2, for sufficient small x, is thermodynamically stable. Such compounds will be able to recreate soluble base during an equilibration. An equilibration is a temperature treatment at sufficient high temperature for a sufficient time in a given atmosphere. During equilibration, surface lithium compounds are reformed at the surface. This requires that lithium diffuses from the bulk to the surface. This obviously would create cationic vacancies which are energetically non-preferred. So oxygen needs to be released from the sample as well, which deletes the cationic vacancies. So the restoration of soluble surface base will happen at a temperature when the bulk oxygen equilibrates.

(25) Hence for a the recreation of surface base 2 mechanisms are required:

(26) 1) Li diffusion from the bulk to the surface (which would create a cationic vacancy), and

(27) 2) local rearrangement of cations and diffusion of oxygen, including the release of oxygen to the atmosphere, which process annihilates the cationic vacancy.

(28) For such processes to occur at a reasonable rate a minimum temperature is required. Obviously process 1) (Li diffusion) happens already at room temperature (otherwise the cathode could not work at room temperature in a battery). Process 2) involves an oxygen equilibration which typically happens at temperatures above 400-500 C.

(29) Treatment temperature: it is advantageous to establish the optimum treatment temperature of the pristine coating layer. If the temperature is too low then surface base is not sufficiently decomposed. If the temperature is too high then surface base is restored as explained above, which continuously reacts with the glassy coating. At the optimum treatment temperature a part of the glassy coating has reacted with the surface base. If the temperature is too high all glassy coating has reacted with lithium and the fully lithiated crystalline phase (fx. Li.sub.2SiO.sub.3) forms.

(30) Why does the base increase at higher temperature? At higher T Li from the bulk continuously replaces surface lithium which again reacts with the coating layer. This process will continue until the glassy coating layer is fully lithiated (and usually not anymore glass phase). Then the process will continue to restore surface base until the equilibrium is reached, the surface containing the equilibrium soluble base and additional the lithium in the fully lithiated coating layer. These contributions add up to a value which is larger than the base content of an uncoated reference sample.

(31) Coating thickness: the glassy coating may be thick enough so that it can decompose the Li containing surface base without itself becoming fully lithiated. If the glassy surface coating is too thick then the cathode performance deteriorates because of the low conductivity and a lower content of electrochemically active material. In the case of Li polysilicate Li.sub.2Si.sub.5O.sub.11 in an embodiment the coating level is between 0.1 to 0.6 mol % of Li.sub.2Si.sub.5O.sub.11 per mol LiMO.sub.2 (this corresponds from about 1000 ppm to 1% by weight silicon).

(32) The invention is further illustrated in the following examples:

Example 1: Preparation of a Pristine Glass-Coated Cathode Powder

(33) This example illustrates the preparation of cathode powders, being free of sulfate impurity and having a pristine (i.e. non-heat treated) glassy coating. As coated LiMO.sub.2 precursor an example cathode material LiMO.sub.2 with M=Ni.sub.0.5Mn.sub.0.3Co.sub.0.2 (or 532 compound) and an average particle size of 10 m is used. The precursor is prepared from a blend of mixed metal hydroxide MOOH and Li.sub.2CO.sub.3 and has a Li:M ratio of about 1.05. Firing is performed in air at 930 C. for 10 h. The MOOH is prepared by precipitation of a metal sulfate solution with NaOH and NH.sub.4OH solution. The MOOH has a tap density of about 1.8 g/cm.sup.3. Such mixed metal hydroxides typically contain 0.3-0.6 wt % of sulfate impurity. The LiMO.sub.2 precursor contains 0.53 wt % of sulfate, being in the form of Li.sub.2SO.sub.4 salt impurity. As it is desired to investigate the glassy coating without cross contamination by the Li.sub.2SO.sub.4, the LiMO.sub.2 precursor is first washed in water, followed by drying. This treatment removes most of the sulfur, resulting in a low sulfate impurity of 0.041 wt %. Table 1 illustrates the preparation process, and FIG. 1 shows a SEM micrograph of the precursor after removing sulfate impurity by washing.

(34) TABLE-US-00001 TABLE 1 Preparation of samples Preparation EX0296 Origin: Blend of MOOH + Li.sub.2CO.sub.3, Li:M 1.05 Firing: air, 930 C., 10 h EX0306 Origin: EX0296 Washing (1 kg/1 L water), filtering drying

(35) A glassy coating is achieved by a treatment which is called slurry doping. A suitable amount of water (approx. 300 ml/kg), containing an appropriate amount of dissolved Li.sub.2Si.sub.5O.sub.11 is added to 2 kg of the precursor, resulting in a slurry of high viscosity. Different samples are prepared containing 0, 0.03, 0.1 and 0.3 mol % Li.sub.2Si.sub.5O.sub.11 per 1 mol of LiMO.sub.2, respectively. After stirring the slurry is dried at 150 C. in air, followed by sieving. Practically all dried powder is recovered after sieving, so the final silicate content of the cathode powder will be near to the target value. During drying most of the Li.sub.2Si.sub.5O.sub.11 precipitates as a thin glassy film onto the surface of the particles, including the pores and gaps between grains. In this way a precursor powder is achieved which has (1) a low sulfate impurity and (2) is coated by thin layer of glass.

Example 2: Preparation of a Series of Final Test Samples

(36) A series of final test samples is prepared by heat treatment in air of the pristine glass-coated cathode powder sample of Example 1. The heat treatment temperature varies between 200 and 600 C.; treatment time is 5 h. Sample size is 150 g. The content of soluble base is measured by pH titration. Coin cells are prepared, the discharge capacity and irreversible capacity of the first cycle between 4.3-3.0 V are measured, whereafter the samples are tested under harsh conditions: cycling between 4.5-3.0V, charge and discharge after cycle 3 at 1 C rate (1 C=180 mA/g). Of interest is a) the reversible capacity and b) the cycle stability (fade rate as %/100 cycles). Table 2 lists the obtained results, i.e. the performance as a function of Li.sub.2Si.sub.5O.sub.11 coating level and heat treatment temperature.

(37) TABLE-US-00002 TABLE 2 0 mol % 0.03 0.1 mol % Base QD Qirr F Base Base QD Qirr F 200 C. 28.90 171.5 9.6 56.0 29.9 27.7 170.5 9.9 69.8 300 C. 30.8 26.9 173.2 9.7 40.4 400 C. 36.70 175.2 9.1 52.7 35.6 27.3 174.0 9.7 26.8 500 C. 56.5 50.6 171.9 10.2 51.7 600 C. 43.8 171.1 11.0 43.2 79.3 94.8 170.6 10.4 57.5 0.3 mol % Base QD Qirr F 200 C. 28.10 168.0 10.8 78.1 300 C. 26.50 169.7 10.6 71.1 400 C. 25.50 171.6 10.5 36.5 500 C. 48.9 169.1 11.1 61.1 600 C. 94.0 166.9 11.8 72.8

(38) (Soluble) Base is in mol/g of cathode, QD is 1.sup.st cycle discharge capacity at 0.1 C (mAh/g) measured between 4.3 and 3.0V, Qirr is the irreversible capacity in % and F is the fade rate during 50 cycles of harsh cycling extrapolated to 100 cycles. Cycling scheme: cycles 1-3 rate performance measurements at resp. 0.1 C, 0.5 C, 1 C with cycling between 3.0-4.3V, Cycles 3-50:1 C charge, 1 C discharge, all cycles 3.0-4.5V.

(39) Obviously, at 400 C. the best performance is achieved. Under all conditions (0, 0.03 mol. 0.1 mol %, 0.3 mol %) the cycle stability (values of Qirr and F) as well as the reversible capacity is at its optimum. At the same time the base content is still small. We also observe that Li.sub.2Si.sub.5O.sub.11 coated samples, after heat treatment to about 300-400 C. show much improved cycle stability (expressed by the fade rate F).

(40) Similar experiments are repeated with different pristine coated cathode powders (being more spherical and having higher tap density). We consistently observe a maximum of capacity and a tow soluble base content, with a much improved cycle stability for 0.05-0.5 mol % Li.sub.2Si.sub.5O.sub.11 coated samples, after a heat treatment at about 400 C.

Example 3: Preparation of a Series of Final Test Samples without Intermediate Wash

(41) An example pristine glass-coated cathode product LiMO.sub.2 (M=Ni.sub.0.5Mn.sub.0.3Co.sub.0.2) is prepared as described in Example 1. However, for preparing the precursor cathode powder a differenthigh densityMOOH is used (tap density>2.0 g/cm.sup.3). In this example no intermediary wash is performed. The sulfur content of the precursor sample is about 0.4 mol %. This example will demonstrate that Li.sub.2Si.sub.5O.sub.11 glassy coated samples, even if sulfur is present, have good cycle stability at high rate, show improved capacity and have reduced content of soluble base.

(42) A Li.sub.2Si.sub.5O.sub.11 coating is applied similar as described in Example 1, with the exception that the cathode precursor powder is not washed. Samples with 2 coating levels, 0.1 and 0.3 mol % are prepared. These coated samples undergo a heat treatment and testing similar as described in Examples 1 & 2, with the exception that only the harsh testing was applied, thus the capacity and irreversible capacity are obtained at 0.1 C rate between 4.5 and 3.0V. Table 3 lists the obtained results: Compared with the reference sample (which is the precursor or core used for the coating) a dramatic reduction of soluble base content is observed. The cycle stability under harsh conditions almost reached that of the non coated sample. The 0.1% coated sample, after heat treatment, shows a clear improved 1.sup.st cycle capacity. Best overall performance is obtained at 0.1 mol % coating level after a heat temperature of 400 C.

(43) TABLE-US-00003 TABLE 3 base content and electrochemical performance obtained for 0.1 and 0.3 mol % Li.sub.2Si.sub.5O.sub.11 coated product, as a function of heat treatment temperature 0.1 mol % 0.3 mol % Base QD Qirr F Base QD Qirr F 200 C. 61.00 192.82 10.47 39.3 60.40 189.24 11.70 64.5 300 C. 56.70 192.24 10.81 26.9 55.20 189.08 12.20 35.0 350 C. 53.90 400 C. 51.40 194.05 11.14 26.6 41.30 191.47 11.90 27.8 450 C. 39.40 500 C. 119.30 190.66 11.71 15.6 108.50 188.51 12.21 33.4 600 C. 170.20 188.64 12.89 19.4 252.70 184.85 13.40 17.9 Uncoated reference (=precursor) 85.80 189.25 12.35 21.3

(44) Cycling scheme: cycles 1-3 rate performance measurements at resp. 0.1 C, 0.5 C, 1 C with cycling between 3.0-4.5V, Cycles 3-50:1 C charge, 1 C discharge, all cycles 3.0-4.5V.

Example 4: Other Examples of Glassy Coating

(45) This example shows that there are other example embodiments of coatings. In this example the washed cathode precursor LiMO.sub.2 with M=Ni.sub.0.5Mn.sub.0.3Co.sub.0.2 of Example 1 is used. Slurry doping is performed in the same manner as in Example 1, with the exception that instead of dissolved Li.sub.2Si.sub.5O.sub.11 other lithium chemicals (in wet solution) are used. The solution used for slurry drying is prepared by dissolve and add stoichiometric controlled amounts of LiOH to a diluted acid solution: Boric acid: H.sub.3BO.sub.3+LiOH.fwdarw.LiBO.sub.2+2H.sub.2O Polyphosphoric acid: HPO.sub.3+LiOH.fwdarw.LiPO.sub.3+H.sub.2O

(46) LiPO.sub.3, obtained from dried LiPO.sub.3 solution is a glass, as well as a Li acceptor: LiPO.sub.3+2LiOH.fwdarw.Li.sub.3PO.sub.4+H.sub.2O. LiBO.sub.2, as it contains boron, can also form a glass.

(47) After slurry coating the samples are dried, and fired for 5 h in air at different temperatures. Final samples are tested for surface area, coin cell performance and content of soluble base. Table 4 lists the preparation conditions and obtained results. The soluble base content is in mol/g of cathode, QD is the 1.sup.st cycle discharge capacity at 0.1 C (mAh/g), and F is the fade rate during 50 cycles of harsh cycling extrapolated to 100 cycles (see cycling scheme below). Table 4 also contains some data of Example 2; but for a coating with 0.1 mol % Li.sub.2Si.sub.5O.sub.11. The table shows that the LiPO.sub.3 and LiBO.sub.2 coatings, at the different treatment temperatures, give a less significant improvement compared to the results of Example 2, whereas Li.sub.2Si.sub.5O.sub.11 coated cathodes, near to 400 C., show a very sharp maximum of capacity and sharp minimum of cycle stability (energy fade rate), and at the same time, have a still lower base content than the reference.

(48) TABLE-US-00004 TABLE 4 Preparation and testing results Heating Added to 1 mol T ( C., QD Qirr Fade Base LiMO.sub.2 air) mAh/g (%) (%/100) mol/g 0.2 mol % 200 C./5 h 170.47 9.74 44.72 36.2 LiPO.sub.3 400 C./5 h 172.77 9.15 40.79 51.6 700 C./5 h 168.91 11.32 47.95 47.5 930 C./5 h 169.21 11.94 35.95 52.5 0.5 mol % 200 C./5 h 171.69 9.55 61.44 64.9 LiBO.sub.2 400 C./5 h 173.00 9.51 53.65 74.7 700 C./5 h 169.72 10.61 51.35 135.6 930 C./5 h 168.86 11.47 41.42 144.5 Reference 200 C./5 h 171.47 9.58 70.38 28.9 H.sub.2O 400 C./5 h 175.16 9.12 61.25 36.7 700 C./5 h 171.07 11.09 52.21 43.8 930 C./5 h 168.58 12.08 47.07 43.6 Example 1 200 C./5 h 170.47 10.12 59.49 27.6 Li.sub.2Si.sub.5O.sub.11 400 C./5 h 171.74 9.64 33.40 32.7 700 C./5 h 167.34 11.31 57.64 134.4 930 C./5 h 167.56 13.14 52.21 88.5

(49) Cycling scheme: cycles 1-3 rate performance measurements at resp. 0.1 C, 0.5 C, 1 C with cycling between 3.0-4.3V, Cycles 3-50:1 C charge, 1 C discharge; all cycles 3.0-4.5V. Fade rate calculated from energy (capacity x average voltage), extrapolated to 100 cycles.

Example 5: Li2Si5O11 Coated Cathodes with Higher Ni Content

(50) When increasing the Ni content in LiNMO we achieve higher capacity but, at the same time, the content of soluble base increases, which is highly disadvantageous for some applications. This example shows that Li.sub.2Si.sub.5O.sub.11 coating allows to reduce significantly the base content of LiMO.sub.2 materials, where M=Ni.sub.0.6Mn.sub.0.2Co.sub.0.2. We will refer to this composition as 622. Here also, for example Li.sub.2Si.sub.5O.sub.11 coated materials, we observe a clear optimum of performance at about 400 C. treatment temperature. Similar as in Examples 1-3 Li.sub.2Si.sub.5O.sub.11 pristine coated cathode materials are prepared from washed (sulfate free) and non-washed precursors by slurry doping, and this is followed by a heat treatment at 200-500 C.

(51) Because a 622 compound has a high base content, the coating content is set at 0.15 mol % Li.sub.2Si.sub.5O.sub.11 per 1 mol LiMO.sub.2. The typical base content of a reference sample is 85-110 mol/g. Compared to this value, Li.sub.2Si.sub.5O.sub.11 coated cathodes obtained from washed LiNMO have 40-50 mol/g, whereas Li.sub.2Si.sub.5O.sub.11 coated cathodes obtained from non-washed LiNMO have about 80 mol/g. Table 5 lists the preparation conditions and obtained results. FIG. 2 shows the properties of Li.sub.2Si.sub.5O.sub.11 coated LiNi.sub.0.6Mn.sub.0.2Co.sub.0.2O.sub.2 as function of treatment temperature: the soluble base content is in mol/g of cathode, Q is the 1.sup.st cycle discharge capacity at 0.1 C (mAh/g), and F is the fade rate during 50 cycles of harsh cycling extrapolated to 100 cycles. In the figure: O: non-washed samples, washed samples. We observe a clear optimum of performance near 400 C. At lower and higher temperature inferior cycle stability is observed. For Li.sub.2Si.sub.5O.sub.11 coated cathodes obtained from non-washed precursor, we also observe a clear minimum of base content.

(52) TABLE-US-00005 TABLE 5 Preparation and testing results Added to 1 mol Heating QD Qirr Fade Base LiMO.sub.2 T ( C., air) mAh/g (%) (%/100) umol/g Non washed 200 C./5 h 197.4 11.0 61.9 85.3 precursor 250 C./5 h 197.1 11.2 59.9 80.6 0.15 mol % 300 C./5 h 199.2 11.2 46.4 77.9 Li.sub.2Si.sub.5O.sub.11 350 C./5 h 201.1 11.1 53.0 78.3 400 C./5 h 200.3 11.2 46.0 81.3 450 C./5 h 199.3 11.2 59.5 105.3 500 C./5 h 193.4 12.0 52.7 151.7 Washed 200 C./5 h 43.7 precursor 250 C./5 h 43.3 0.15 mol % 300 C./5 h 204.0 10.2 80.6 42.4 Li.sub.2Si.sub.5O.sub.11 350 C./5 h 203.8 10.4 74.3 46.7 400 C./5 h 204.8 10.3 63.2 53.3 450 C./5 h 202.3 10.5 86.5 72.4 500 C./5 h 199.8 11.1 101.3 97.6

(53) Cycling scheme: cycles 1-3 rate performance measurements at resp. 0.1 C, 0.5 C, 1 C with cycling between 3.0-4.5V, Cycles 3-50:1 C charge, 1 C discharge, all cycles 3.0-4.5V.

Example 6: XPS and SEM Measurements

(54) This Example describes the investigation of 3 LiMO.sub.2 samples of Example 2, (M=Ni.sub.0.5Mn.sub.0.3Co.sub.0.2) coated with 0.3 mol % Li.sub.2Si.sub.5O.sub.11, and heat treated at 200 C. (EX6.1), 400 C. (EX6.2) and 600 C. (EX6.3), using X-ray Photoelectron Spectroscopy (XPS) and high resolution SEM, to support the findings that at intermediate temperatures (400 C.) there is an optimum in battery performance (lowered base content, improved fade performance).

(55) The experiment is designed to prove that: 1) At too high temperatures (600 C.) there is a strong increase in base content accompanied by a structural change in the pristine Li-silicate coating. 2) At too low temperatures (200 C.) no structural changes occurs in the pristine Li-silicate coating. 3) At intermediate temperatures (400 C.) an optimum is reached in battery performance due to the diffusion of a small amount of Li (originating from the surface) into the silicate layer and the consumption of surface base during this Li-diffusion. 4) The Li-silicate layer forms a continuous overlayer and the grain boundaries are closed when the Li-silicate is heat-treated at 400 C.
XPS Data

(56) The results of the C, Si and Li spectra are summarized in Table 6.

(57) TABLE-US-00006 TABLE 6 Overview of apparent atomic concentrations (at %) measured in the surface layer after deconvolution of the C 1s, Si 2s and Li 1s spectra into their different contributions. Element C 1s Si 2s eV 154.1 152.7 Li- Li- Li/Si ratio poor rich Li 1s (corrected Sub- 284.8 286.5 288.0 289.7 sil- sili- 54.7 for Li in species CH CO CO CO.sub.3 icate cate Li.sup.+ Li.sub.2CO.sub.3) EX6.1 8.4 1.4 <1.0 <1.0 15.0 / 6.0 0.4 EX6.2 10.2 1.3 <1.0 <1.0 14.5 / 7.6 0.5 EX6.3 10.5 <1.0 <1.0 2.5 / 7.2 19.1 2.0
Conclusions for Table 6:
1 C 1s: 1.1 CH, CO and CO are typical contaminations always observed with XPS. 1.2 The CO.sub.3 peak at 289, 7 eV is typical for the presence of Li.sub.2CO.sub.3 surface base. As mentioned in previous examples, other surface base (such as LiOH) are also present but cannot be identified using XPS. 1.3 Small decrease of CO.sub.3 from 200 C. to 400 C. 1.4 Strong increase of CO.sub.3 between 400 C. and 600 C.
2 Si 2s: 2.1 At 200 C. and 400 C., the silicate is present in a Li-poor form (154.1 eV). 2.2 Between 400 C. and 600 C. a transformation occurs into a Li-rich silicate (152.7 eV). 2.3 Strong decrease of Si between 400 C. and 600 C. due to enrichment of the Li-silicate layer with Li.
3 Li 1s: 3.1 Slight diffusion of Li to the silicate-coating from 200 C. to 400 C. 3.2 From 400 C. to 600 C., Li strongly diffuses to the silicate layer at the surface.
4 Li/Si Ratio in the Coating: 4.1 Li/Si is initially equal to the one of Li.sub.2Si.sub.5O.sub.11 (Li/Si=0.4). 4.2 Somewhere between 200 C. and 400 C., the diffusion of Li from the bulk to the surface slowly begins thereby causing an enrichment of the Li-silicate layer with Li (Li/Si ratio increases to 0.5). 4.3 Above 400 C., the Li-silicate coating is heavily enriched with Li from the bulk and forms Li.sub.2SiO.sub.3 (Li/Si increases to 2.0)

(58) The formation of a continuous overlayer of Li-silicate is confirmed by the decrease in Ni and Co signals as compared to uncoated material (Table 7). With increasing temperature, the Li-silicate layer is becoming thicker as can be seen by the further decrease in Ni, Co and O signals.

(59) TABLE-US-00007 TABLE 7 Overview of apparent atomic concentrations (at %) measured in the surface layer after deconvolution of the Ni 2p, Mn 2p, Co 2p and O 1s spectra into their different contributions. Sample Co 2p Mn 2p Ni 2p O 1s eV 780.2 654.0 855.2 / CoO.sub.x MnO.sub.x NiO.sub.x / Uncoated 2.3 3.3 9.7 52.4 EX6.1 1.3 2.2 5.3 58.8 EX6.2 1.0 2.7 4.0 56.7 EX6.3 <1.0 4.3 2.2 51.0

(60) The above XPS data support the following model: 1. At 200 C., the Li-silicate is present as Li.sub.2Si.sub.5O.sub.11 and the surface base level is similar as in a non-coated product. 2. Between 200 C. and 400 C., a small amount of Li (originating from surface base) diffuses into the silicate layer thereby causing a slight enrichment of the silicate layer and a small decrease in surface base. 3. Between 400 C. and 600 C., the diffusion of Li continues strongly. The silicate becomes very Li-rich. In this Li-rich form, the silicate can act as a CO.sub.2-sorber thereby forming Li.sub.2CO.sub.3 and increasing the surface base.
SEM Data

(61) SEM data (FIGS. 3.1, 3.2 and 3.3) show that, for all temperatures, the Li-silicate coating forms a continuous overlayer. At 200 C., strong aggregates of Li-silicate are visible and the grain boundaries are open. At higher temperatures (400 C. and 600 C.) the aggregates seem to have melted away and the grain boundaries are clearly closed.

Example 7: Full Cell Testing of LiSi Coated Cathodes

(62) This example demonstrates that excellent results are obtained with example LiSi coated cathode material incorporated in real full cells. These cells are winded pouch type with approx. 800 mAh capacity. In the cells 0.1 mol Li.sub.2Si.sub.5O.sub.11 coated LiMO.sub.2 cathode materialswith M=Ni.sub.0.5Mn.sub.0.3Co.sub.0.2are tested.

(63) The example shows results for the following samples, originating from the same mixed MOOH (M=Ni.sub.0.5Mn.sub.0.3Co.sub.0.2): 1) EX7.1: Normal reference sample prepared from the low density MOOH of Example 1, fired at 955 C. using a Li:M ratio of 1.035 (without intermediate wash) 2) EX7.2: Washed reference sample, refired at 600 C., 5 h 3) EX7.3: Washed sample, 0.1 mol Li.sub.2Si.sub.5O.sub.11 coated, refired at 400 C., 5 h 4) EX7.4: Mass Production (MP) line LiMO.sub.2 with M=Ni.sub.0.5Mn.sub.0.3Co.sub.0.2 prepared from high density MOOH as used in Example 3 5) EX7.5: 0.1 mol % Li.sub.2Si.sub.5O.sub.11 coated MP line LiMO.sub.2 (no intermediary wash, heat treatment at 400 C.)

(64) The results of full cell testing are summarized in Table 8. Bulging is measured by inserting the fully charged cell into an oven, heating within 1 h to 90 C., and measuring the thickness by a suitable thickness gauge directly attached to the cell. Bulging generally depends on the electrolyte chosen, here we use a standard EC/DEC electrolyte which has not been optimized to achieve tow bulging.

(65) TABLE-US-00008 TABLE 8 Full cell testing of 0.1 mol Li.sub.2Si.sub.5O.sub.11 coated LiMO.sub.2 (M = Ni.sub.0.5Mn.sub.0.3Co.sub.0.2) compared to references Capacity after Capacity after Bulging (%) 4.2 V 500 cycles (%) 350 cycles (%) 4 h (15 h) @ 90 C. at 25 C. at 45 C. EX7.1 64.2 (75.5) 71.4 74.5 EX7.2 38.5 (53.7) 87.1 68.7 EX7.3 18.8 (30.2) 88.2 80.9 EX7.4 19.7 (30.8) 82.0 68 EX7.5 14.3 (25.4) 90.7 92.3

(66) The results show that the high temperature performance of 0.1 mol Li.sub.2Si.sub.5O.sub.11 coated samples is much better (less bulging, better cycle stability) than that of both references. Cycle stability at room temperature is better as well. Other properties (rate performance, capacity, safety) are similar or slightly better.

(67) The results show that an intermediary wash is not necessary. The 0.1 mol % Li.sub.2Si.sub.5O.sub.11 coated sample EX7.5 has the best cycle stability at room temperature as well as at 45 C. and it shows the least bulging. Other properties (rate performance, capacity, safety) are similar or slightly better than the references.

Example 8: Method to Prepare LiSi Coated Cathodes

(68) This example illustrates an example method to produce LiSi coated cathode material. The method is easy scalable to mass production scale. In Examples 1-3 a good performance is achieved by the slurry doping method. Cathode precursor powder is immersed in an aqueous Li.sub.2Si.sub.5O.sub.11 solution to obtain a slurry of relatively high viscosity. In this way a good penetration of the solution into the pores and a 100% surface coverage is achieved. About 300 ml are used for 1 kg product. The slurry is subsequently dried, followed by soft grinding and sieving. However drying of such a slurry at industrial scale requires capital investment and energy, hence it is not very cheap. Therefore it is preferred to investigate if instead of slurry doping a coating can be applied using less water.

(69) Commercial LiMO.sub.2 (M=Ni.sub.0.5Mn.sub.0.3Co.sub.0.2) is obtained from a mass production line. 1.7 kg of the LiMO.sub.2 powder is immersed into a heated 5 L reactor, and smaller amounts of higher concentrated Li.sub.2Si.sub.5O.sub.11 solution are added during soft agitation. The total amount of Li.sub.2Si.sub.5O.sub.11 is fixed at 0.1 mol % per 1 mol LiMO.sub.2. Ideally the solution is sprayed into the agitated powder. The reactor is connected to a vacuum pump, and during continuous agitation at 80 C., the powder is dried. This process can easily be scaled up to mass production level.

(70) It is observed that the processing becomes easierthe powder becomes less stickyas smaller amounts of higher concentrated solution is added. The upper limit of the concentration of the solution is attained when the surface coverage of the glass solution deteriorates, especially within open pores. Only if the glassy solution has excellent surface wetting properties good results are expected.

(71) After drying, the powder is heat treated at 400 C. 4 samples are prepared using 400, 300, 200 and 100 ml of solution, for 1.7 kg LiMO.sub.2, but respecting the concentration of Li.sub.2Si.sub.5O.sub.11 of 0.1 mol % per 1 mol LiMO.sub.2 (meaning that the 100 ml solution is 4 more concentrated than the 400 ml). The best results are obtained with 100 ml. Table 9 lists the results for reversible capacity and base content. The results are compared with a reference sample which previously has been prepared at a small scale level, using the slurry doping method for the same mass production precursor. With only 60 ml/kg LiMO.sub.2 solution similar results as the reference are obtained (a difference of 1 mAh/g capacity and 2 mol/g base are within experimental scattering). The data clearly demonstrate that smaller amounts of higher concentrated glassy solution can have excellent surface wetting properties, so that the coating process at large scale easily can be performed by using small amounts of higher concentrated glass solution. We speculate that the excellent wetting properties are related to the solution being a dissolved glass (forming a glass after drying) and not a crystalline salt.

(72) TABLE-US-00009 TABLE 9 Properties of coated LiMO.sub.2 as function of dilution of glassy solution QD base Sample Preparation mAh/g Qirr % mol/g EX0542 Slurry doping (300 ml/kg) 171.0 11.7% 37.60 reference EX0775 400 ml (235 ml/kg) 168.2 12.3% 42.18 EX0776 300 ml (180 ml/kg) 168.9 12.3% 41.19 EX0777 200 ml (120 ml/kg) 168.9 12.1% 45.27 EX0778 100 ml (<60 ml/kg) 169.8 11.9% 36.04

Example 9: Xray Analysis Confirms Li Accepting Property of Glass Phase

(73) This example displays the Li accepting properties of Li polysilicate glass. The Li accepting character is strong enough to decompose LiOH and Li.sub.2CO.sub.3, and hence Li polysilicate glass is strong enough to decompose the Li containing surface base.

Experiment 9A

(74) Li.sub.2Si.sub.5O.sub.11 liquid glass is dried at 200 C. FIG. 4 shows the result of the X-ray diffraction analysis, indicating that at 200 C. an amorphous glass is obtained. A single very broad peak at about 23 degree is typical for Si-based glass. Thus, coating a cathode material with liquid glass will result, after drying, in a glassy coating. We believe that the glassy coating very well covers the surface of the cathode powder.

(75) Alternatively liquid glass is dried at 400 and 600 C. At these temperatures Li.sub.2Si.sub.5O.sub.11 disproportionates into a Li-rich crystalline Li.sub.2SiO.sub.3 phase and an amorphous glass phase. Almost all sharp peaks are indexed as Li.sub.2SiO.sub.3 (PDF 01-070-0330, space group Cmc21). A few minor remaining peaks at 24.85, 23.8 and 37.6 can be indexed as Li.sub.2Si.sub.2O.sub.5. The position of the glass phase peak moves left towards 21.5 degree. Obviously the glass phase has a lower Li:Si ratio than 2:5.

(76) It can be concluded that, according to temperature, (1) Li.sub.2Si.sub.5O.sub.11 coating followed by drying results in a glassy coating; (2) Li.sub.2Si.sub.5O.sub.11 disproportionates into Li.sub.2SiO.sub.3 and a low lithium glassy phase, which is also a Li acceptor; (3) As long as not all glass has reacted with lithium, even at 600 C. a glassy coating remains, (4) No Li.sub.2CO.sub.3 is formed, despite that the drying is performed in air, which contains enough CO.sub.2 to form Li.sub.2CO.sub.3.

Experiment 9B

(77) This experiment illustrates that the Li accepting properties of Li.sub.2Si.sub.5O.sub.11 are strong enough to decompose Li.sub.2CO.sub.3. Li.sub.2Si.sub.5O.sub.11 liquid glass is dried at 120 C. The glass is grinded and mixed with Li.sub.2CO.sub.3 (10 g glass and 4 g Li.sub.2CO.sub.3). At 450 C. Li.sub.2CO.sub.3 is not very reactive, so to enhance the contact a pellet is pressed and fired at 450 C. for 72 hours in air.

(78) Assuming that Li.sub.2CO.sub.3 reacts completely with Li.sub.2Si.sub.5O.sub.11 to form Li.sub.2SiO.sub.3 (reaction scheme: Li.sub.2Si.sub.5O.sub.11+4Li.sub.2CO.sub.3.fwdarw.5Li.sub.2SiO.sub.3+4CO.sub.2) requires a mass ratio Li.sub.2CO.sub.3:glass of approx. 0.9:1. In this experiment we use a mass ratio of 1:2.5, meaning that there is a large excess of glass. FIG. 5 shows the results of the X-ray diffraction analysis: the top graph is the XRD pattern for the mixture Li.sub.2CO.sub.3:glass, the bottom graph is the XRD pattern for the same mixtures after heating at 450 C. for 72 hr. The glass phase changes (the broad hump moved towards smaller angle and becomes slightly more narrow (12-35 to about 10-30). Diffraction peaks of the Li.sub.2CO.sub.3 phase, as well as the intensity of the broad hump (glass phase) clearly decreases, indicating that more than 50% of the Li.sub.2CO.sub.3 has been decomposed, partially consuming the glass phase. Additionally Li.sub.2SiO.sub.3 peaks (with higher intensity ratio of Li.sub.2SiO.sub.3 to glass phase than in Experiment 9A) form. The X-ray diffraction pattern clearly proves that Li.sub.2Si.sub.5O.sub.11 glass is a Li acceptor, strong enough to decompose Li.sub.2CO.sub.3. In the case of excess Li.sub.2Si.sub.5O.sub.11 at 450 C. a phase mixture forms which contains remaining (but modified) glass and Li.sub.2SiO.sub.3, where some of the Li.sub.2SiO.sub.3 originates from a decomposition of Li.sub.2CO.sub.3 by the glass. The results of the XRD analysis are consistent to the observations by XPS of Example 6 (at about 400 C. the coating layer hascompared to the initial Li.sub.2Si.sub.5O.sub.11an increased Li content by having decomposed the surface Li.sub.2CO.sub.3).