LITHIUM MANGANESE OXIDE AND METHODS OF MANUFACTURE
TECHNICAL FIELD
This invention relates to metal doped lithium manganese oxide and methods of its production.
BACKGROUND
Given the huge commercial success of rechargeable lithium batteries there is considerable interest in the development of improved intercalation compounds for use as the electrodes in such batteries. Appropriate compounds must possess a sufficient cap fftity for providing energy, must have appropriate electric conductivity and, preferably, should be stable at elevated temperatures. Furthermore, they must be able to be produced in commercial quantities at a viable cost.
The trivalent manganese compound LiMn02 has been identified as a possible cathode material for lithium ion batteries. In particular, monoclinic (m-) LiMn02 (space group C2/m), having a layered rock salt structure similar to LiCo02 and LiNi02, exhibits promising electrochemical properties. However. Li_vln02 obtained via conventional solid state reaction usually has an orthorhombic (o-LiMn02) structure (space group Pmnm). since m-LiMnOi is not usually stable under these synthesis conditions.
To date, m-LiMn02 has been obtained by an ion exchange reaction of lithium salts with NaMn02 as precursor, or by a mixed alkaline-hydrothermal reaction (Armstrong & Bruce, Nature 38 . 499 (1996); Tabuchi et. al. J. Electrochem. Soc.145. L49 (1998)..
International patent application no. PCT US97/18839, by Massachusetts
Institute of Technology, discloses compositions having a general formula LixMyNx02,
and including the compound LiAlo.25Mno.75O2, being a layered monoclinic structure. The added aluminium apparently stabilises the layered monoclinic modification and allows direct synthesis of the compound. Chiang et al., in Electrochem. Solid State Lett. I, 13 (1998) describe the preparation of both LiAlo._5Mno.75O2 and LiAlo.05Mno.95O2 in the layered monoclinic structure. However, the synthesis route used to prepare the compounds in the above references (PCT US97/18839 and Chiang et al.) is based on the coprecipitation of the precursor hydroxides out of aqueous nitrate solution and subsequent freeze drying of the obtained hydroxide mixture. The dry mixture is then fired at 945°C in an inert atmosphere.
This synthesis method is complex and expensive and the nitrates of aluminium and manganese may be carcinogens and mutagens. Thus, the method may not be appropriate for commercial production.
Davidson et al. in United States patent No. 5,370,949 and J. Power Sources
54, 205 (1995) disclose compounds of composition Li2CrxMn2.x04 prepared by high temperature solid state reactions of lithium carbonate with chromium (III) oxide and manganese (IV) oxide, or of LiCr02 with o-LiMn02. Compounds for which x > 1.5 apparently had a layered hexagonal structure similar to LiCr02. Compounds for which 0.1 < x <1.25 gave X-ray diffraction patterns which could be indexed to a tetragonal cell based on a structure like the lithiated spinel-related phase Li2Mn2θ4. Dahn et al. in J. Electrochem. Soc. 145. 851 (1998) showed that the data described by Davidson et al. in US 5,370,949 and J. Power Sources 54, 205 (1995) are better explained by a monoclinic layered crystal structure.
The data published by Davidson et al. and Dahn et al, if taken in combination, therefore demonstrate that chromium can be used in a solid state reaction to stabilise a layered monoclinic structure for LiCrxMnι-x02 for x > 0.05 (dividing the subscripts in the formulae given by Davidson et al. and Dahn et al. by a factor of 2 for consistency with other descriptions).
The selection of element or elements doped into LiMnC>2 and the proportions of dopant may be important in determining the extent to which a stable LiMn02 composition may be produced with the required electrochemical properties by a commercially viable process.
Moreover, a combination of dopants used to stabilise a layered monoclinic phase of LiMn02 may facilitate the synthesis procedure and/or provide electrochemical advantages over the prior art compounds using only aluminium or chromium as the dopant.
OBJECTS OF THE INVENTION
It is therefore an object of the present invention to provide a method of synthesising lithium manganese oxide intercalation compounds which reduces or overcomes the abovementioned problems, or which at least provides the public with a useful alternative.
It is a further and/or alternative object of the present invention to provide lithium manganese oxide intercalation compounds having improved characteristics, or at least to provide the public with useful alternative compounds.
Other objects of the present invention may become apparent from the following description which is given by way of example only.
SUMMARY OF THE INVENTION
According to one aspect of the present invention there is provided a metal-doped lithium manganese oxide composition having the formula LiM1 miM2 m2...Mx mxMn-O2 wherein M1, M2...MX are each different metal ions, ml+m2+...mx+z is substantially equal to 1, ml+m2+...mx is greater than zero but less than 0.75 and z and at least mi are each greater than zero, providing that where m2...mx are zero M1 is Cr and 0<ml<0.05 or M1 is Ga.
Preferably, M1, M2...MX are each selected to have an ionic radius in their relevant oxidation state, similar to or less than that of Mn3+.
Preferably, M1, M~...MX are each selected from the group consisting of Al, Ga, Cr, Co, Fe, B, V, Ti and Ru.
Preferably, M1 may be Al.
Preferably, M2may be selected from Cr, Co, Fe, B, V, Ti and Ru and is most preferably Cr.
Preferably, M1, M2...M each has the property of existing in a minimally distorted octahedral sphere of co-ordination.
Preferably, M1, M"...Mx each has a formal oxidation state of 3+.
Preferably, mι+m2+...mx may be greater than zero but less than about 0.25.
Preferably the composition is in predominantly monoclinic form.
According to a further aspect of the invention there is provided a method of synthesising a metal-doped lithium manganese oxide composition having the formula LiM'mi M2 m2...Mx mx Mnz 02 wherein M1, M2...M are each different metal ions,
m l+m2+...mx + z is substantially equal to 1, ml+m2+...mx is greater than zero but less than 0.75 and z and at least ml are each greater than zero, providing that where m2...mx are zero and M1 is Cr then 0<ml<0.05, in stable form by solid state reaction.
Preferably, the solid state reaction includes:
mixing reagents including a lithium compound, a manganese compound and at least one additional metal compound, each compound selected from an oxide, a hydroxide and/or a carbonate, and
heating the mixture in a reactor under inert atmosphere, to a temperature exceeding 800°C for more than 2 hours.
Preferably, the lithium compound may be lithium carbonate.
Preferably, the manganese compound may be manganese oxide.
Preferably, the or each metal of the at least one additional metal compound may be selected to have an ionic radius in its relevant oxidation state, similar to or less than that of Mn3+.
Preferably, the or each metal of the at least one additional metal compound may be selected from the group consisting of Al, Ga, Cr, Co, Fe, B, V, Ti and Ru.
Preferably, the at least one additional metal compound includes aluminium oxide. Preferably, a transitional alumina.
Preferably, the at least one additional metal compound alternatively, or in addition, includes chromium oxide.
Preferably, the method may further include ball milling the mixed reagents.
Preferably, the mixture may be heated to a temperature in the range substantially 900°C to 1080°C. More preferably substantially 1000°C to 1080°C for a period of at least 5 hours.
Preferably, the inert atmosphere may have a sufficiently low p02 to facilitate the formation of manganese in its 3+ oxidation state.
Preferably, the p02 may be substantially 10"5 atm or less.
Preferably, the mixture may be heated in a reactor having a high temperature stable lining.
According to a further aspect of the invention there is provided a method of synthesising a metal-doped lithium manganese oxide composition of the formula LiMI mιM2 m2...Mx mxMnzθ2 wherein M1, M2...Mx are each different metal ions, ml+m2+...mx+z is substantially equal to 1, ml+m2+...mx is greater than zero but less than 0.75 and z and at least ml are each greater than zero, in stable predominantly monoclinic form, the method including the steps of:
- mixing reagents including a lithium compound, a manganese compound and at least one additional metal compound, each compound selected from an oxide, a hydroxide and/or a carbonate,
heating the mixture in air to a temperature > 200°C for at least 5 hours to produce oxygen-rich lithium manganese oxides,
further heating the mixture in a reactor under inert atmosphere to a temperature exceeding 800°C for at least 2 hours.
Preferably the heating in air may be at a temperature in the range 600 to 850°C. Most preferably substantially 700°C.
Preferably the heating under inert atmosphere may be at a temperature of substantially 1000 to 1080°C.
According to a further aspect of the present invention there is provided a method of synthesising a metal doped lithium manganese oxide composition substantially as herein described and with reference to the accompanying examples.
Other aspects of the present invention may become apparent from the following description which is given by way of example only and with reference to the accompanying examples.
BRIEF DESCRIPTION OF FIGURES
Figure 1; Effect of temperature on LiMn02 production by high temperature solid state reaction with metal doping in a steel reactor in accordance with Example I.
Figure 2: Effect of reaction time on LiMn02 production in accordance with Example I.
Figures 3 A-F: X-ray diffraction patterns showing the effects of temperature on the production and phase of LiMn02 from a high temperature solid state reaction with metal doping in an alumina-lined reactor in accordance with Example II:
• - o-LiMn02
+ - m-LiMn02 o - Li2Mn03
Mn304
Figures 4 A-D: X-ray diffraction patterns showing the effects of chromium doping on the production and phase of LiMn02 from a high temperature solid state reaction with metal doping in an alumina-lined reactor, in accordance with Example IV:
• - o-LiMn02
+ - m-LiMn02 o - o-Li2Mnθ3
Mn304
LiCrQ2
Figure 5: X-ray diffraction patterns of LiCrxMnι-xθ2 for x= 0.01, 0.03, 0.05, 0.10 and 0.20, prepared according to Example V (+ = m-LiMn02, o = o-LiMn02).
Figure 6: X-ray diffraction pattern of LiAlo.05Mno.95O2 prepared according to Example VI (+ = m-LiMnC>2, o = o- LiMnQ2).
Figure 7: Schematic representation of the cell construction as used in Examples VI, VII, VIII and X.
Figure 8: X-ray diffraction pattern of LiAlo.10Mno.90O2 prepared according to Example VII (o = o-LiMn02, all other peaks m-LiMn02).
Figures 9A-D: X-ray diffraction patterns of (A) LiAlo.05Cro.01Mno.94O2,
(B) LiAlo.05Cro.03Mno.9_O;>, (C) LiAlo.05Cro.05Mno.90O2,
and (D) LiAlo.02Cro.05Mno.93O2 prepared according to Example VIII (o = o-LiMn02).
Figures 10A&B: X-ray diffraction patterns of (A) Lio.5Cro.05Mno.95O2 and (B) LiCro.05Mno.95O2 prepared according to Example IX.
Figures 11A&B: X-ray diffraction patterns of LiAlo.05Mno.95O2 prepared according to Example X (A) after air step (* = spinel, # = Li2Mn03), (B) after nitrogen step (o = o-LiMn0 ).
Figures 12A&B: X-ray diffraction patterns of (A) Lio.5Gao.10Mno. 0O2 and (B) LiGao.10Mno.90O2 prepared according to Example XI (o = o-LiMnθ2).
Figures 13A-D: X-ray diffraction patterns of (A) Lio.5Cro.05Gao.05Mno.90O2, (B) LiCro.05Gao.05Mno.90O2, (C) Lio.5Cro.05Bo.05Mno.90O2, and (D) LiCro.05Bo.05Mno.90O2 prepared according to Example XII (o = o-LiMnθ2).
DETAILED DESCRIPTION OF THE INVENTION
Investigations were carried out to identify new metal-doped LiMnθ2 compositions and provide a robust, simple and commercially viable solid state synthesis of aluminium, chromium and/or other metal doped layered monoclinic m- LiMnθ2, which could have commercial significance as a cathode in lithium ion batteries.
The basic reaction underlying the synthesis method and compositions of the present invention involves the conversion of, for example, lithium carbonate and
manganese oxide to lithium manganese oxide. One possible reaction scheme would be:
Li2C03 -> Li20 + C02
The invention involves methods of synthesising stable forms of metal-doped LiMn02, in layered monoclinic form, by solid state reaction. The invention also relates to specific stable compositions formed by these methods.
Preferred, stable compositions were produced with the starting reagents lithium carbonate (Li2C03), manganese (IV) oxide (Mn02) and aluminium oxide (AI2O3) heated in a reactor under inert atmosphere to a temperature in the range 800°C to 1080°C for a period of at least 2 hours. The formation of LiMn02 in monoclinic form, may be facilitated by raising the temperature to between about 950°C and 1080°C, and using a transitional alumina as the source of aluminium.
Other preferred, stable compositions were produced with the starting reagents lithium carbonate (L_2C03), manganese (IV) oxide (Mnθ2), aluminium oxide (A1203) and chromium oxide (Cr203) heated in a reactor under inert atmosphere to a temperature in the range 800°C to 1080°C for a period of at least 2 hours.
Other preferred, stable compositions were produced in a two-stage reaction by firstly reacting lithium carbonate (Li2C03) and manganese carbonate (MnCC ) with one or more of aluminium oxide (A1203), chromium oxide (Cr20 ), gallium oxide (Ga2θ3) and boric acid (H3B03) in air at a temperature of 600-850 °C to produce oxygen-rich lithium manganese oxides, followed by a second heating step under nitrogen at 800-1080 °C.
In this two-stage reaction, if the lithium, manganese and dopant reagents in the first stage are combined in the appropriate ratios to give a product with a
Iithium/(manganese + dopant) molar ratio of 1, the reaction in air will produce a two phase mixture of a spinel phase LixMn2. θ4 and rock salt phase Li2Mn03. Both phases may contain the dopant elements in their crystalline structure. When this product is heated under N2 flow, the two phases react together to give doped LiMn02 according to the reaction scheme:
LiM2χMn2.2x04 + Li2M Mnι- θ3-^3 LiMxMnι-x02 + ' _ 02
where M may be one or more dopant elements.
Alternatively, the reagents in the first stage can be combined in the appropriate ratios to give a product with a lithium/(manganese + dopant) ratio of approximately 0.5. The product of the reaction in air will then be a single-phase doped spinel oxide. This can then be combined with additional Li reactant in the second stage under N2 to give doped LiMn02, e.g.:
LiM2XMn2.2x04 + '/a L_2C03→2 LiMxMnι.x02 + lΛ CO: + V_ 02
An advantage of the two-stage method is that the dopant precursor compounds are reacted with the manganese precursor compound in air, which provides more reactive conditions for the decomposition of some metal salts than an inert atmosphere such as N2. This may allow the selection of metal salts as starting compounds which may be difficult to react together to give a single-phase product under an inert atmosphere. Spinel LiMn2θ4 can be prepared with a large number of different dopants at a range of concentrations. Many of these doped LiMn204 compounds are known and have been well characterised in terms of the effects of the dopant element on the crystal structure. This allows for ready identification of the doped spinel intermediate phase by techniques such as X-ray powder diffraction. The relatively facile formation of doped spinel phases in air then provides an atomic-level dispersion of the dopant element in the manganese oxide lattice before proceeding to the conversion step under N2. The conversion step under N2 therefore involves
essentially lithium diffusion into the spinel phase from the Li2Mn03 phase or additional Li reactant, some reduction of Mn4+ to Mn3+, and an accompanying crystal phase transformation.
It will be appreciated by those skilled in the art that different lithium, manganese and dopant compounds may also be employed as the starting reagents, but generally selected from carbonates, oxides and hydroxides.
It is considered that the metal ions most preferred for doping of the LiMn02 should be selected by several criteria, firstly ionic radius, and secondly appropriate charge state. The incorporation of a metal ion with an ionic radius similar to or less than that of Mn3+is likely to mitigate the distortion of the octahedral geometry around the Mn3+. Thus, preferably, the metal ion or ions should have an ionic radius, in their relevant oxidation state, similar to or less than that of Mn ". Hence, as an alternative to, or in addition to aluminium, the dopants may be selected from, for example, B, Ga, Cr, Co, Fe, V, Ti and Ru. Preferably, the metal ion or ions may also have the property of existing in a minimally distorted octahedral sphere of co-ordination.
Example I
Exact stoichiometric amounts (based on LiAl0 o5Mn095θ2) of lithium carbonate (99.9% pure), manganese (IV) oxide (90+% pure) and aluminium oxide (99.8% pure) were weighed out, mixed and subsequently homogenised for 30 minutes using a vibrating table. These mixtures were finally fired under inert atmosphere (BOC, Zero Grade Nitrogen <5ppm 02, flow rate =
5Lmin"') in a proprietary stainless steel rotary reactor fitted inside a muffle furnace. After reaction the reactor was removed from the muffle furnace and cooled to <50°C by blowing air across the reactor body whilst maintaining the inert atmosphere. The duration of the described cooling process was approximately 20 minutes.
To evaluate the temperature effect precursor mixtures were fired at temperatures between 700°C and 1120°C. The reaction time was usually between 3 and 5 hours. The produced samples were analysed using X-Ray diffraction (XRD) to assess phase purity. The monoclinic polymorph of LiMn02 started to form at 1000°C. The other phases identified by XRD were orthorhombic 0-LiMnO2, tetragonal LiA102, and rock salt Li2Mn03. The latter, and the spinel phase LiMn20 , were preferentially formed if the oxygen partial pressure in the reactor exceeded lO^atm. Thus, the inert atmosphere should preferably involve a pθ2 of 10"5 atm or less. This is sufficient to facilitate the formation of manganese in its 3+ oxidation state.
The effect of temperature is summarised in Figure 1.
The highest amounts of the monoclinic polymorph m-LiMn02 were obtained at 1060°C, yielding 85% m-LiMn02, 5% o-LiMn02, 5% of LiA102, and 5%
Li Mn03.
It should be noted that this reaction was performed with in-situ grinding by means of a steel rod.
Higher temperature (above 1060°C) appeared to reduce the amount of the monoclinic phase.
More detailed temperature effect results, and X-ray diffraction data are included below under Example II.
Samples of the end product were digested in hydrochloric acid and hydrogen peroxide and subsequently analysed by AA (for Li) and ICP-OES (for Mn and Al). The molar ratio Li:Al:Mn was found to be substantially 1.00:0.05:0.95+ 5%.
Reaction time was investigated at a constant reaction temperature of 1060°C.
The results are shown in Figure 2.
The reaction seems to be completed in approximately three hours. Reaction times of longer than 5 hours usually lead to a decrease of the LiMn02-phases and the formation of spinel phase. Reaction time and temperature are likely to be dependent on particle size, and the required reaction may be achieved in a shorter time employing reagents, particularly manganese oxide, having a smaller particle size.
Example II
In the following example an alumina-lined steel reactor was employed in place of a conventional rotary stainless steel reactor.
Exact stoichiometric amounts (based on LiAlo.05Mno.95O2) of lithium carbonate (99.9% pure), manganese (IV) oxide (90% + to <99.0% pure) and aluminium oxide (99.8% pure) were homogenised for 30 minutes using a vibrating table. The resulting mixtures were fired at temperatures between 800°C and 1100°C under inert atmosphere (BOC, zero grade nitrogen) in alumina-lined stainless steel rotary reactors fitted inside a muffle furnace and cooled to less than 50°C by blowing air across the reactor body whilst maintaining the inert atmosphere.
Samples resulting from processing for about 5 hours at 1040°C had average phase compositions of approximately 80% o-LiMn02 and 10% m-LiMn02.
X-ray diffraction patterns (see Figure 3 A-F) showed a clear shift in the proportion of LiMn02 at a temperature of 950°C (compare Figures 3C and D) and higher, and clear identification of the monoclinic form at a temperature of 1050°C (Figure 3F). However, as can be seen from Table 1, the proportion of
phases was unexpected in comparison with the results of Example I using a stainless steel reactor. Even at 1050°C there was onlv about 15%) m-LiMnO?.
Table 1: Temperature effect on LiMnO2 proportions and phases, all at 5 hours reaction time
Elemental analysis revealed the presence of high amounts of iron and chromium in samples produced in the conventional steel reactors in comparison with the samples produced in alumina-lined reactors. The Fe content of samples prepared in a steel reactor was generally in the range 1- 3%, with Cr in the range 0.05-0.2%. These figures were reduced by about a factor of 10 in the alumina reactor.
It was concluded that a few percent of one or both of Fe and Cr, transferred into the reaction mixture from the steel reactor, aided the formation of the monoclinic LiMn02 phase under reaction conditions where it is otherwise difficult to form m-LiMnO_.
Example HI
The process of Example II was repeated in alumina boats in a tube furnace, with the addition of chromium (in the form of Cr203) to the reaction mixture and subsequent firing at temperatures from 1050 to 1100°C for 3 to 7 hours, under an inert atmosphere. As in Example II, the mixture was first heated to
about 200°C for about an hour before increasing the temperature to the selected level. The results are shown in Table 2.
Table 2: Chromium doping of Aluminium-doped LiMnO2
As can be seen from Table 2, the presence of Cr203 1 to 2.5 wt%, with a reaction temperature of 1050°C for a period of at least 5 hours, resulted in a marked change in the phase composition of the end product in comparison with the results obtained under similar conditions, but without chromium, in the alumina-lined reactor.
Whilst these results were obtained using chromium oxide, it is anticipated that other metals (such as Co and Fe) or combinations of metals would similarly facilitate the formation of Al-doped m-LiMnθ2- It is also considered that the chromium, or an alternative metal, may reduce the temperature at which Al- doped m-LiMn02 is formed.
Example IV
The effects of chromium doping were investigated in an alumina-lined reactor using the reagents described for Example II with the addition of chromium oxide (Cr203), and at 1050°C for 5 hours.
X-ray diffraction patterns for compositions resulting from the use of different starting proportions of Cr203 are shown in Figures 4A - D. X-ray diffraction analysis revealed the estimates shown in Table 3 for the content of the compositions.
Table 3: Chromium doping of aluminium-doped LiMnO2, in alumina reactor
At 3 wt% Cr203, LiMn02 was still predominantly in the orthorhombic form (45%), although there was a 35% proportion of monoclinic form contrasting with only 15% under the same conditions but without chromium doping (see Figure 3). With 5 wt% Cr2θ3, LiMnC>2 was predominantly in the monoclinic form (70%o). The higher proportion of chromium oxide required in this Example compared with Example III could indicate inadequate mixing of the starting materials. Pregrinding of the reaction mixture could significantly improve the effect of chromium.
Example V
The effects of doping chromium alone in small quantities in LiMn(_>2 were systematically investigated in a vertical cylindrical reactor lined with a high
1.
temperature stable lining. Exact stoichiometric amounts of lithium carbonate (99.9% pure), Mn02 (99.0% pure) and Cr203 (99.0% pure) to give products of formula LiCr.Mnι..O_, where = 0.01, 0.03, 0.05, 0.10, and 0.20, were mixed and ball-milled for 24 hours in a rotating shaker. The resulting mixtures were loaded into the bottom of the reactor and fired at 1000 °C for 5 hours with a flow of zero grade nitrogen percolating through the reaction mixture from an inlet at the bottom of the reactor, with stirring at regular intervals to ensure homogeneity. Elemental analyses of the products using the method described in Example VI confirmed that the end products had the correct Li:Cr:Mn ratios.
The X-ray diffraction patterns of the products are shown in Figure 5. The phase mixtures obtained, estimated from the ^-ray diffraction analyses of the products, are given in Table 4. When x is between 0.05 and 0.20 the products are 100% single phase monoclinic LiCr_Mni_.0 , as indicated by the data of Davidson et al. in J. Power Sources 54, 205 (1995) and Dahn et al. in J. Electrochem. Soc. 145, 851 (1998). However the results show that predominantly monoclinic LiCrJV__ii..0 can also be obtained when x=0.01 or 0.03, a result not predicted by the aforementioned publications.
Table 4: Chromium doping of LiMn02 at 1000 °C
The monoclinic layered structure (space group C2/m) of the LiCr.Mn1._O2 products was confirmed by Rietveld analysis of the X-ray diffraction data. The unit cell parameters calculated from the X-ray diffraction data by the Rietveld refinement method are shown in Table 5 for the single-phase monoclinic LiCr.Mn1-.O2 products where ..=0.05, 0.10 and 0.20, compared
with published data for non-doped monoclinic layered LiMnOi (Capitaine et al., Solid State Ionics 89, 197 (1996)).
Chromium doping results in changes in the crystal unit cell dimensions consistent with substitution of Cr ions into Mn sites of the crystal structure. In LiMnθ2, the extent of the monoclinic crystal distortion resulting from the Jahn Teller distortion of the octahedral coordination environment around the Mn ions is reflected in the magnitude of the ratio between the a and b crystal axes. The systematic decrease in the alb ratio with increasing amounts of Cr in the crystal is an effect of a minimally distorted octahedral sphere of coordination around the Cr ions. The atomic positions of the cations in the crystal lattice were also refined using the Rietveld method. The results confirmed that Cr and Mn ions share the transition metal sites in the layered crystal structure. No Mn or Cr ions were detected in the interlayer (lithium) sites of the crystal.
Table 5: Unit cell parameters for the single-phase monoclinic LiCrΛrMnι..ϊO2 products
Capitaine et al. (Solid State Ionics 89, 197 (1996)).
Example VI
The effects of using a transitional alumina as the source of Al in the solid state reaction were investigated in a vertical cylindrical reactor with a high
temperature stable lining. Exact stoichiometric amounts of lithium carbonate (99.9% pure), Mn02 (99.0% pure) and theta-Al203 (99.95% pure) to give a product of formula LiAl0 osMno 95O1 were mixed and ball-milled for 30 minutes. The resulting mixture was loaded into the bottom of the reactor and heated to 1000°C for 3.5 hours with a flow of zero grade nitrogen percolating through the reaction mixture from an inlet at the bottom of the reactor, with stirring at regular intervals to ensure homogeneity. After 3.5 hours the product was allowed to cool in the furnace over a period of 15 hours, maintaining the nitrogen flow until the temperature dropped below 50 °C.
The X-ray diffraction pattern of the product, shown in Figure 6, shows that the product contains approximately 90% m-LiMn02. and approximately 10% o-LiMn02. No additional impurity phases were detected.
Samples of the product were digested in hydrochloric acid and hydrogen peroxide and analysed by AA (for Li) and ICP-OES (for other metal ions). The molar ratio Li:Al:Mn was found to be 1.03:0.05:0.94, consistent with the expected stoichiometry within experimental error (±5%) of the analytical technique. Less than 0.1% Cr, 0.1%o Ni, and 0.4% Fe were detected in the product, indicating that contamination from the reactor lining was negligible and unlikely to have contributed to the formation of the monoclinic phase.
The formation of a predominantly monoclinic structure at 1000 °C in this case is attributed mainly to the higher reactivity of the theta-Al203 by comparison with the aluminium-containing reactants used in the preceding examples.
When the reaction was carried out under the same conditions but using an alpha-Al 03 as the source of aluminium, mainly orthorhombic LiMnθ2 was produced. When alpha- A1203 was used, XRD also detected traces of LiA10 in the product. This was not observed when the theta-Al2θ3 was used, suggesting that a more complete reaction of the A120 with the manganese
oxide is the critical factor favouring the formation of monoclinic phase under these particular conditions.
The unit cell parameters for the monoclinic phase, obtained by performing a simultaneous Rietveld refinement of the crystal parameters for both the monoclinic and the orthorhombic phase from the X-ray diffraction data, are compared with those given by Capitaine et al. (Solid State Ionics 89, 197 (1996)) for non-doped monoclinic LiMn02 in Table 6. The monoclinic LiAlo.05Mno.95O2 phase of this example shows a contracted crystal lattice by comparison with the undoped compound. This is consistent with the presence of the Al ion, which has a smaller ionic radius than trivalent Mn, sharing the same crystallographic site as Mn. The contraction is more significant along the a crystal axis than the b crystal axis, consistent with a minimally distorted octahedral sphere of co-ordination around the Al ions since the magnitude of the alb ratio is related to the average degree of distortion of the octahedron of oxygen ions around the cations in the Mn layer.
Table 6: Unit cell parameters for the monoclinic LiAlo.05Mno.95O2 phase of Example VI, compared with reported values for undoped monoclinic LiMnO2 (Capitaine et al.)
These results demonstrate that the choice of alumina reagent and ball milling of the reactant mixture before heating can influence the temperature at which a predominantly monoclinic form of aluminium doped LiMn02 is obtained using the solid state reaction method of the invention.
Electrodes were prepared by mixing 80 wt.%> of LiAlo.05Mno.95O2, 12 wt.% acetylene black and 8 wt.% poly(vinylidene fluoride) as a slurry in 1-methyl-
2-pyrrolidinone (NMP). The slurry was coated onto aluminium foil. After evaporation of the solvent, the coating was pressed on the aluminium foil and
annealed at 150 °C under vacuum. Circular electrodes measuring 14 mm in diameter were then punched from the coated foil. The circular electrodes were weighed individually and the active mass (the total weight of the circular electrode multiplied by the fraction of the electrode weight made up by LiAlo 05Mn095O2) calculated. The electrodes were then dried at 150 °C under vacuum to remove traces of water and transferred to an argon-filled dry glove box (<1 ppm water).
The electrodes were assembled into electrochemical cells within the argon- filled glove box using 2032 button cell hardware. The cell assembly is schematically illustrated in Figure 7. The electrode containing the LiAlo 05Mn095O2 was the cathode 1. The anode 2 was a circular disk of lithium foil having a thickness of 0.38 mm, pressed into a stainless steel lid 3. A polymeric gasket 4 was then positioned over the lip of the lid. A porous glass fibre disk separator 5 wetted with 1M LiPF6 in (50 wt.% ethylene carbonate + 50 wt.%) dimethylene carbonate) electrolyte solution was placed between the anode and cathode. A stainless steel disk 6 and spring 7 were then positioned behind the cathode and the entire assembly hermetically closed within the stainless steel casing 8 by crimp sealing.
Cells prepared by this procedure were cycled more than 200 times between 2.0 and 4.4 V at ambient temperature and elevated temperature (55 °C). A constant charging current of 30 or 75 mA/g was applied until 4.4 V was reached, then the cell was held at 4.4 V until the current dropped below 3 mA/g. Cells were discharged at constant currents of 30 or 75 mA/g. Typical discharge capacities obtained by this procedure at the 30 mA g discharge rate are given for the 1st, 16th, and 200th cycles in Table 8 in Example VIII. The discharge capacities at ambient temperature increase with cycling. At 55 °C the discharge capacities are particularly high and show good stability.
Example VII
Exact stoichiometric amounts of lithium carbonate (99.9% pure), Mn02 (99.0% pure) and theta-Al203 (99.95% pure) to give a product of formula LiAlo.10Mno.90O2 were mixed and ball-milled for 30 minutes. The resulting mixture was loaded into the bottom of the reactor described in Example VI and heated to 1050°C for 10 hours under a flow of zero grade nitrogen. After 10 hours the product was allowed to cool in the furnace over a period of 15 hours, maintaining the nitrogen flow until the temperature dropped below 50°C.
The X-ray diffraction pattern of the product, shown in Figure 8, shows that the product contains approximately 95% m-LiMn02, with the remainder being o-LiMn02 phase. No additional impurity phases were detected.
The unit cell parameters of the monoclinic phase, calculated from the XRD pattern, are a = 5.416(2) A, b = 2.804(1) A, c = 5.382(2) A, and β= 115.91°. Like the monoclinic LiAlo.o5Mn0.95θ2 phase of Example VI, the LiAl0. ιoMno.9o0 material of this example shows a contracted crystal lattice by comparison with the undoped compound.
Electrodes and electrochemical cells containing the LiAl0.ιoMιio.9o02 material were prepared as described in Example VI, and the cells were charged and discharged following the same procedures. Typical discharge capacities obtained at the 30 mA/g discharge rate are given for the 1st, 16th, and 200th cycles at ambient and 55 °C in Table 8 in Example VIII. The material shows slightly lower initial capacities than the LiAlo.05Mno. O2 compound of Example VI, but the capacity increases with cycling to values which are similar to those of the LiAlo.o5Mn0.95θ2 material.
Example VHI
Compositions LiAl CrvMnι. -yθ2 were prepared by reacting lithium carbonate (99.9% pure), Mn02 (99.0% pure), theta-Al203 (99.95% pure) and Cr203 (99.0% pure). Exact stoichiometric amounts of the reactants to give the compositions LiAlo o5Cr0oiMn09 O2, LiAl0 osCro 03Mn09202, LiAlo osCro osMno 9o02. and LiAlo o2Cr0 osMno 930 were mixed and ball-milled for 24 hours in a rotating shaker, then loaded into the vertical cylindrical reactor of Example VI and heated at 1050 °C for 5 hours under a flow of zero grade nitrogen.
The X-ray diffraction patterns of the products are shown in Figure 9(A-D). In the case of the compositions LiAloo5Cro oιMn094θ2 and LiAl0 osCro 03M1109_02, the pro^μcts were 80-90 % layered monoclinic phase, with the remaining phase being o-LiMn02. No other impurity phases were observed. The compositions LiAl0 osCro os no 90O2 and LiAl0 o_Cr0 o<Mn093θ2 were phase- pure layered monoclinic materials. Cell parameters for these phases are given in Table 7.
Table 7: Unit cell parameters for the monoclinic
products
Electrodes and electrochemical cells containing the LiAlo
05Cr
0 o
5Mn
09o0
2, and LiAlo o
2Cr
0 osMno 9 O
2 materials were prepared as described in Example VI, and the cells were charged and discharged following the same procedures. Typical discharge capacities obtained at the 30 mA/g discharge rate are given for the 1
st, 16
th, and 200
th cycles at ambient temperature and 55 °C in Table 8. The cells cycled at 55 °C showed similar excellent capacity retention over 200 cycles as the LiAl
0 osMno 95O
2 and LiAlo ιoMn
09o0
2 materials of Examples VI and VII. However the LiAl
0 osCro osMno 90O2, and LiAl
0 o_Cr
0 osMn
093O
2 materials showed increased overall capacity at both 55 °C and ambient temperature when compared with the LiAl
0 osMno 95O
2 and LiAl
0 ιoMn
09oθ
2. materials.
The co bination of chromium plus aluminium together in monoclinic layered LiMn02 therefore provides a material with improved cycling characteristics over monoclinic layered LiMn02 doped with aluminium alone. The addition of chromium to the aluminium-doped LiMn02 also facilitated the obtaining of a single-phase monoclinic layered material.
Table 8: Discharge capacities for the monoclinic LiAlϊMnι-.02 materials of Examples VI, VII & IX, and the LiAlvCr,,Mn1-γ-v02 materials of Example VHI.
Exact stoichiometric amounts of lithium carbonate (99.9%) pure), manganese carbonate (99.5% pure) and Cr203 (99.0% pure) to give a product of formula Lio.5Cro.05Mno.95O2 were mixed and ground in an acetone slurry using a mortar and pestle. The mixture was loaded into a vessel formed from an inert metal alloy and fired in a furnace at 700 °C for 20 hours in air. The X-ray diffraction pattern of the product, in Figure 10A, shows that it was a single phase cubic spinel with a unit cell parameter of a = 8.235 A. The X-ray diffraction data are consistent with doping of the chromium on the manganese site of spinel LiMn204.
This compound was ground in acetone with an equivalent of lithium carbonate to give a product of formula LiCro.05Mno.95O2, and heated under a flow of zero grade nitrogen at 1000 °C for 18 hours. The X-ray diffraction pattern of this product, in Figure 10B, shows formation of phase-pure monoclinic layered LiMn02, with unit cell parameters a - 5.431 A, b = 2.807A, c = 5.382 A, and β = 1 15.94 °. These cell parameters are very close to those obtained for the LiCro.05Mno.95O2 compound of Example V.
This example shows that a two stage solid state process, in which a doped spinel phase is formed by reaction in air and then reacted with additional lithium under nitrogen, can be used to synthesise doped LiMn02 materials. This reaction route provides an alternative to the single-step reaction, and may be useful when employing starting oxides or salts (as the source of lithium, manganese or dopant) which are difficult to react together to give a single- phase product under an inert atmosphere. Such oxides or salts may react more easily in air to give a doped oxidised lithium manganese oxide product, which can then be converted into the doped LiMn02 in inert atmosphere in a second stage.
It will be appreciated by those skilled in the art that doped oxidised lithium manganese oxide compounds can be formed by solid state reactions at temperatures ranging from 200 °C up to 900 °C, and that a range of temperature conditions could therefore be used in the first stage of this process. The second, conversion step in inert atmosphere should preferably be performed at a higher temperature than the first step.
Example X
The two stage reaction process was used to prepare Al-doped LiMn02. Exact stoichiometric amounts of lithium carbonate (99.9% pure), Mnθ2 (99.0% pure) and theta-Al θ3 (99.95%) pure) to give a product of formula LiAlo.05Mno.95O2 were mixed and ball-milled for 30 minutes. The resulting mixture was heated at 700°C in air for 8 hours, cooled, then fired at 1000 °C for 5 hours under a flow of zero grade nitrogen. The product was cooled over a period of one hour, maintaining the nitrogen flow until the temperature dropped below 50 °C.
X-ray diffraction of the product after the first stage of the reaction in air (Figure 11 A) showed a mixture of cubic spinel phase and Li2Mn03. The presence of L_2Mnθ3 in this case is due to the amount of lithium, one mole of lithium per mole of manganese + aluminium, being too high for a single- phase spinel. The X-ray diffraction pattern of the product after the second stage under nitrogen (Figure 1 IB) shows complete conversion of the spinel and L_2Mnθ3 phases to LiMnθ2 product. The product is approximately 90% monoclinic layered LiMn02, with around 10% o-LiMn02, practically identical to the LiAlo.05Mno.95O2 product of Example VI.
Electrodes and electrochemical cells containing the LiAlo.05Mno.95O2 material of this example were prepared as described in Example V, and the cells were charged and discharged following the same procedures. Typical discharge capacities obtained at the 30 mA/g discharge rate are given for the 1st, 16th,
and 200th cycles at ambient and 55 °C in Table 8 in Example VIII. The electrochemical characteristics are almost identical to those of the LiAlo.05Mno.95O2 product of Example VI.
Example XI
The two-stage synthesis procedure described in Example IX was used to prepare a compound of stoichiometry LiGao.ιoMn0.9o02. Exact stoichiomefric amounts of lithium carbonate (99.9% pure), manganese carbonate (99.5%> pure) and gallium oxide Ga2θ3 (99.9% pure) to give a product of formula
Lio.5Gao.10Mno.90O1 were mixed and ground in an acetone slurry using a mortar and pestle. The mixture was loaded into a vessel formed from an inert metal alloy and fired in a furnace at 700 "G^for 20 hours in air. The X-ray diffraction pattern of the product, in Figure 12 A, shows that it was a single phase cubic spinel with a unit cell parameter of a = 8.227 A.
This compound was ground in acetone with an equivalent of lithium carbonate to give a product of formula LiGao.ιoMno.9o02, and heated under a flow of zero grade nitrogen at 1000 °C for 15 hours. The X-ray diffraction pattern of this product, in Figure 12B, shows 80% formation of monoclinic layered LiMn02, with unit cell parameters a = 5.430 A, b = 2.810 A, c = 5.387 A, and β= 115.78 °. The remaining 20% of the product is o-LiMn02.
This result demonstrates that gallium may be used as an alternative, or in addition, to aluminium or chromium as a means of stabilising the layered monoclinic modification of LiMn02.
Example XH
The two- stage synthesis procedure described in Example IX was used to prepare compounds of stoichiometry LiCr
0 osGao osMn
09o0
2 and LiCr
0 05B
0 osMno
9o0
2. Exact stoichiometric amounts of lithium carbonate (99.9% pure), manganese carbonate (99.5% pure), Cr
20
3 (99.9% pure) and gallium oxide Ga 0
3 (99.9% pure) or boric acid H3BO3 (99.9%> purity) to give products of formulae Li
0 sCro osGao osMno 9o0
2 and Li
05Cr
00
5B0 osMno
90O
2 were mixed and ground in an acetone slurry using a mortar and pestle. The mixtures were loaded into a vessel formed from an inert metal alloy and fired in a furnace at 700 °C for 20 hours in air. The X-ray diffraction patterns of the products in Figure 13A&C show that both were single phase cubic spinels with a unit cell parameter of a = 8.227 A for the Li
0 sCro osGao osMno
90O
2 material and a = 8.234 A for the
osMn
0 0O
2 material.
These compounds were each ground in acetone with an equivalent of lithium carbonate to give products of formulae LiCr0 osGao osMn090O2 and LiCr005B0 osMno 90O2, and heated under a flow of zero grade nitrogen at 1000°C for 15 hours. The X-ray diffraction patterns of the products in Figure 13B&D show the formation of phase-pure monoclinic layered LiMnθ2 in both cases, with unit cell parameters a = 5.410 A, b = 2.815 A, c = 5.378 A, and β= 1 15.79 ° for the LiCr0 osGao osMn09o02 material, and a = 5.429 k, b = 2.811 A, c = 5.380 A, and β= 115.86 ° for the LiCr005B005Mn09o02 material.
The above examples show that monoclinic layered LiMn02 may be formed during a high temperature solid state reaction involving metal doping, under specific conditions of time and temperature, and with appropriate starting reagents. Zero grade nitrogen is sufficient to control the manganese oxidation state in the formation of the monoclinic LiMnθ2 material, as only very little Li2Mn03 and no spinel phases were usually formed in the reactions under zero grade nitrogen.
The temperature window for producing layered monoclinic LiMn02 is in the range 800 to 1080°C.
The formation of aluminium doped layered monoclinic LiMn02 from lithium carbonate, manganese oxide and aluminium oxide (or hydroxide) precursors by high temperature solid state reaction is facilitated by the inclusion of an additional metal, such as chromium.
It will be appreciated by those skilled in the art that the selection of reagents, in particular their average particle sizes and reactivity, will influence the optimum conditions of time and/or temperature to achieve the preferred result. As shown in the examples, under appropriately selected conditions, the high temperature solid state reaction method of the invention can produce stable layered monoclinic LiMn02 doped with aluminium, chromium, gallium, or combinations of aluminium, chromium, gallium and boron. Combinations of aluminium and chromium result in improved electrochemical properties over materials doped with aluminium alone. It is anticipated that other combinations may also result in improvements.
Where in the foregoing description reference has been made to specific components or integers of the invention having known equivalents then such equivalents are herein incorporated as if individually set forth.
Although this invention has been described by way of example and with reference to possible embodiments thereof it is to be understood that modifications or improvements may be made thereto without departing from the scope or spirit of the invention.