CA2686496A1 - Improved lithium ion electrochemical cells - Google Patents

Abstract

An electrochemical cell has a negative electrode comprising a liquid crystal templated mesoporous material capable of forming a lithium insertion alloy, and having a relatively high porosity of from 38% to 80%.

Description

IMPROVED LITHIUM ION ELECTROCHEMICAL CELLS

The present invention relates to improvements in the construction of lithium ion electrochemical cells, including capacitors, supercapacitors and batteries, by means of an improved negative electrode (anode) comprising a mesoporous material that is active for lithium insertion.

The mesoporous materials used in the present invention are sometimes referred to as "nanoporous". However, since the prefix "nano" strictly means 10-9, and the pores in such materials may range in size from 10-8 to 10-9 m, it is better to refer to them, as we do here, as "mesoporous".

Although, strictly speaking, the term "battery" means an arrangement of two or more cells, it is used here with its common meaning of a device for storing and releasing electrical energy, whether it comprises one or several cells.

The drive towards `convergence' in electronics, i.e. increasing the functionality of devices such as mobile phones and personal digital assistants (PDAs) has increased the demand for energy and power placed on batteries. At present, we believe that the greatest scope for capacity (energy) improvement lies in the development of the negative electrode. The majority of commercial lithium ion batteries currently use negative electrodes based on carbon. The charge storage capacity of carbon is typically in the region of 300 mAh/g.

Alternatives to carbon are materials which are capable of forming alloys with lithium at low potentials, such as tin, silicon and aluminium. These materials have charge storage capacities up to 2000 mAh/g. However, insertion of lithium into these materials is accompanied by significant expansion of the structure. This causes rapid mechanical breakdown of the material and manifests in cell performance as poor cycle life. In addition, expansion of the electrode material during charging can cause expansion of the entire battery, leading to other perfonnance and safety concerns. As such, the commercial realisation of these high capacity materials has to date been limited and carbon electrodes remain the dominant technology.

In Chemical Communications, 1999, 4, 331-332, J. R. Owen discloses a lithium ion battery negative electrode consisting of an electrodeposited tin film made using a liquid crystal templating route. The paper states: "It would be expected that extensive mesoporosity would significantly reduce internal stresses during expansion and thus decrease the mechanical degradation of the electrodes". Cycle life was found to be poor, however.

We have now surprisingly found that, by engineering higher than normal levels of porosity into liquid crystal templated materials, these materials are better able to cope with expansion on lithiation and so can provide superior cycle life and reduced overall particle expansion. Alternatively, higher capacities may be achieved since the higher porosity allows a greater degree of lithiation (expansion) before mechanical breakdown of the material begins. Particle expansion is reduced since the increased porosity allows more of the expansion to be accommodated by the internal mesopores of the material rather than outward expansion of the particle.

Porosities typically achieved in liquid crystal templating are in the range of approximately 13% to 27%. Porosities in excess of 35% would be considered unusually high, and the usual porosity is around 23%. However, we have found that unexpected benefits may be achieved by using significantly higher porosities.

Thus, the present invention consists in an electrochemical cell comprising a positive electrode, a negative electrode and a non-aqueous electrolyte, where the negative electrode comprises a liquid crystal templated mesoporous material capable of forming a lithium insertion alloy, characterised in that the liquid crystal templated mesoporous material has a porosity of from 38% to 80%.

The invention is illustrated by the accompanying drawing, in which: Figure 1 compares the cycle life behaviour of two cells; one utilising an anode having a porosity of 51 % and made as described in Example 3 and another that used the same construction, the only difference being that 39% porous copper-tin material (made as described in Example 2, using a liquid crystal template) was used.

The preparation and use of liquid crystalline phases is disclosed in US
Patents No 6,503,382 and 6,203,925, the disclosures of which are incorporated herein by reference.

The electrochemical cell of the present invention may be a capacitor, supercapacitor or battery. Where it is a battery, this is normally a secondary, i.e.
rechargeable, battery.

The material capable of forming a lithium insertion alloy may be an element (a metal or metalloid) or it may be a mixture or alloy of one or more elements capable of forming a lithium insertion alloy with one or more elements which cannot form such an insertion alloy or a mixture or alloy of two or more elements each capable of forming a lithium insertion alloy. Examples of elements that are active for lithium insertion are aluminium, silicon, magnesium, tin, bismuth, lead and antimony. Copper is inactive for lithium insertion, but alloys of copper with an element, such as tin, which is active may themselves be active. Other inactive elements include nickel, cobalt and iron.
There is an advantage in including these inactive alloying elements in that their presence effectively dilutes the active material so that less expansion occurs on cycling, leading to further improved cycle life. The preferred active element is tin, and this is most preferably used as an alloy with an inactive element, most preferably copper. -It is a crucial aspect of the present invention that the porosity of the material active for lithium insertion should be from 38% to 80%. The porosity herein is calculated from nitrogen porosimetry (BET) measurements. In general, we have found that cycle life improves as porosity increases. However, too high a porosity will lead to a reduction in the amount of active material present and so may detract from cell performance. Preferably the porosity is in the range from 42% to 75%, more preferably from 44% to 70%. Most preferably the porosity is from 50% to 65%.

The material active for lithium insertion is unlikely to have sufficient mechanical strength on its own to serve as an electrode and, accordingly, it is preferably used in the electrochemical cell on a support, which may also function as a current collector. The support material is thus preferably electrically conductive and preferably has sufficient mechanical strength to remain intact when formed into a film which is as thin as possible. Suitable materials for use as the support include copper, nickel and cobalt, of which copper is preferred both for its cost and its electrical conductivity.

In order to enhance the conductivity of the electrode, the porous material is preferably mixed with an electrically conductive powder, for example: carbon, preferably in the form of graphite, amorphous carbon, or acetylene black;
nickel; or cobalt. If necessary, it may also be mixed with a binder, such as ethylene propylene diene monomer (EPDM), styrene butadiene rubber (SBR), carboxy methyl cellulose (CMC), polyvinyl diene fluoride (PVDF), polytetrafluoroethylene (PTFE), polyvinyl acetate or a mixture of any two or more thereof. The porous material, electrically conductive powder and optionally the binder may be mixed with an organic solvent, such as hexane, water, cyclohexane, heptane, hexane, or N-methylpyrrolidone, and the resulting paste applied to the support, after which the organic solvent is removed by evaporation, leaving a mixture of the porous material and the electrically conductive powder and optionally the binder.

The electrochemical cell also contains a positive electrode. This may be any material capable of use as a positive electrode in a lithium ion cell.
Examples of such materials include LiCoO2, LiMnO2, LiNiCoO2, or LiNiAlCoO2. Like the negative electrode, this is preferably on a support, e.g. of aluminium, copper, tin or gold, preferably aluminium.

The electrolyte likewise may be any conventional such material, for example lithium hexafluorophosphate, lithium tetraborate, lithium perchlorate, or lithium -hexafluoroarsenate, in a suitable solvent, e.g. ethylene carbonate, diethylene carbonate, dimethyl carbonate, propylene carbonate, or a mixture of any two or more thereof.
The cell may also contain a conventional separator, for example a microporous polypropylene or polyethylene membrane, porous glass fibre tissue or a combination of polypropylene and polyethylene.

Preparation of the mesoporous material used as the negative electrode in the cells of the present invention may be by any known liquid crystal templating method.

For example, a liquid crystalline mixture is formed and a mesoporous material is caused to deposit from it. A variety of methods can be used to effect this deposition, including electrodeposition, electroless deposition, or chemical deposition. Of course, to some extent, the method of deposition used will depend on the nature of the material to be 5 deposited.

For example, one method of preparing the mesoporous material comprises electrodepositing material onto a porous support from a mixture comprising at least one source of said material, an organic directing agent and a solvent; by passing charge through said mixture until sufficient of said material has been deposited to form a mesoporous layer on said porous support; and then removing the organic directing agent to produce a mesoporous layer preferably having a substantially regular pore structure and uniform pore size within the desired range, e.g. from 2.5 to 50 nm, on said porous support.

The nature of the source material used in the mixture will depend on the nature of the material to be produced. For example, if the desired material is tin, then a compound of tin, e.g. SnBF4, ,SnCH3SO3, SnC14, or SnC12, should be used. If it is desired to produce a mixture of two or more elements, e.g. an active and an inactive element, then a mixture of the compounds of the respective elements should be used.
Examples of source materials for inactive elements include CuBF4, CuSO4, CuC12, or CoC12.

The organic structure-directing agent is included in the mixture in order to impart a homogeneous lyotropic liquid crystalline phase to the mixture. The liquid crystalline phase is thought to function as a structure-directing medium or template for deposition of the mesoporous layer. By controlling the nanostructure of the lyotropic liquid crystalline phase, mesoporous material may be synthesised having a corresponding nanostructure. For example, porous materials formed from normal topology hexagonal phases will have a system of pores disposed on a hexagonal lattice, whereas porous materials formed from normal topology cubic phases will have a system of pores disposed in cubic topology. Similarly, porous materials having a lamellar nanostructure may be deposited from lamellar phases. Accordingly, by exploiting the rich lyotropic polymorphism exhibited by liquid crystalline phases, liquid crystal technology allows precise control over the structure of the porous materials and enables the synthesis of well-defined porous materials having a long range spatially and orientationally periodic distribution of uniformly sized pores.

Any suitable amphiphilic organic compound or compounds capable of forming a homogeneous lyotropic liquid crystalline phase may be used as structure-directing agent, either low molar mass or polymeric. These may include compounds sometimes referred to as organic directing agents. In order to provide the necessary homogeneous liquid crystalline phase, the amphiphilic compound will generally be used at an high concentration, typically at least about 10% by weight, preferably at least 20%
by weight, and more preferably at least 30% by weight, based on the total weight of the solvent, source material and amphiphilic compound.

Preferably, the organic structure-directing agent comprises an organic surfactant compound of the formula RQ wherein R represents a linear or branched alkyl, aryl, aralkyl or alkylaryl group having from 6 to about 60 carbon atoms, preferably from 12 to 18 carbon atoms, and Q represents a group selected from: [O(CH2)m]õOH
wherein m is an integer from 1 to about 4 and preferably m is 2, and n is an integer from 2 to about 60, preferably from 4 to 12; nitrogen bonded to at least one group selected from alkyl having at least 4 carbon atoms, aryl, aralkyl and alkylaryl; and phosphorus or sulphur bonded to at least 2 oxygen atoms. Other suitable structure-directing agents include monoglycerides, phospholipids and glycolipids.

Other suitable compounds include surface-active organic compounds of the formula R1RZQ wherein Rl and R2 represent aryl or alkyl groups having from 6 to about 36 carbon atoms or combinations thereof, and Q represents a group selected from: -(OC2H4) õOH, wherein n is an integer from about 2 to about 20; nitrogen bonded to at least two groups selected from alkyl having at least 4 carbon atoms, and aryl;
and phosphorus or sulphur bonded to at least 4 oxygen atoms.

Preferably non-ionic surfactants such as octaethylene glycol monododecyl ether (C 12E08 , wherein EO represents ethylene oxide) and octaethylene glycol monohexadecyl ether (C16EO8) or commercial products containing mixtures of related molecules are used as organic structure-directing agents. Other preferred organic directing agents include polyoxyalkylene derivatives of propylene glycol, such as those sold under the trade mark "Pluronic", and ionic surfactants such as CTAB.

It has been found that the pore size of the porous material can be varied by altering the hydrocarbon chain length of the surfactant used as structure-directing agent, or by supplementing the surfactant by an hydrocarbon additive. For example, shorter-chain surfactants will tend to direct the formation of smaller-sized pores whereas longer-chain surfactants tend to give rise to larger-sized pores. The addition of a hydrophobic hydrocarbon additive such as n-heptane, to supplement the surfactant used as structure-directing agent, will tend to increase the pore size, relative to the pore size achieved by that surfactant in the absence of the additive. Also, the hydrocarbon additive may be used to alter the phase structure of the liquid crystalline phase in order to control the corresponding regular structure of the porous metal. By a suitable combination of these methods, it is possible to control the pore size very precisely and over a wide range, extending to much smaller pore sizes (of the order of 1 nm) than could be achieved hitherto.

The solvent is included in the mixture in order to dissolve the source material and to form a liquid crystalline phase in conjunction with the organic structure-directing agent, thereby to provide a medium for deposition of the mesoporous material.
Generally, water will be used as the preferred solvent. However, in certain cases it may be desirable or necessary to carry out the deposition in a non-aqueous environment. In these circumstances a suitable organic solvent may be used, for example formamide or ethylene glycol.

In most cases, the source material will dissolve in the solvent domains of the liquid crystalline phase, but in certain cases the source material may be such that it will dissolve in the hydrophobic domains of the phase.

The mixture may optionally further include a hydrophobic hydrocarbon additive to modify the pore diameter of the porous metal, as explained more fully above.
Suitable hydrocarbon additives include n-heptane, n-tetradecane and mesitylene. The hydrocarbon additive may be present in the mixture in a molar ratio to the structure-directing agent in the range of 0.1 to 4, preferably 0.5 to 1.

Alternatively, the material of which the mesoporous layer is formed may be deposited by electroless deposition. The procedure used to fabricate material by electroless deposition is essentially the same as that used in chemical deposition, described below. The essential difference is that, prior to application of the liquid crystal template to a support, the support is sensitised with a metal salt in order to promote deposition of the mesoporous material only on the support surface rather than throughout the liquid crystal. In summary, the reduction of a metal salt to a metal is facilitated by an appropriate reducing agent just as in chemical deposition.
The presence of the sensitiser confines this deposition to the support surface. A
suitable sensitiser is tin (II) chloride.

As a further alternative, the material of which the mesoporous material is formed may be a metal or other material capable of deposition by reduction or other chemical reaction. In this case, the mixture comprises a source material for the metal or other element, dissolved in a solvent, and a sufficient amount of an organic structure-directing agent to provide an homogeneous lyotropic liquid crystalline phase for the mixture.
Examples of suitable source materials include compounds of the element which are capable of reduction to the element, for example, tin methanesulphonate, copper sulphate, SnBF4, SnC14, SnC12, CuBF4, CuC12. The nature of the solvent is not critical, and is usually aqueo .us.

One or more source materials may be used in the mixture, for reduction to one or more metals. Thus, by appropriate selection of source material, the composition of the porous metal can be controlled as desired. Suitable metals include those described above in relation to the electrodeposition method.

A reducing agent is used to reduce the mixture. Suitable ireducing agents include metals (such as zinc, iron or magnesium), sodium hypophosphite, dimethyl borane, hydrogen gas, and hydrazine, preferably sodium hypophosphite or dimethyl borane.

The structure directing agents and solvents used in this embodiment may be any of those described above in relation to the electrodeposition method.

Typically, the pH of the mixture may be adjusted to a value in the range from to 12. The temperature is generally maintained in the range from 15 to 100 C, preferably 18 to 80 C, more preferably 20 to 60 C.

The mixture and reducing agent are left to stand for a sufficient period to precipitate the porous material, typically overnight at room temperature.
Depending on the nature of the reactants, the mixture may be left for a period of from 15 minutes to 4 weeks, and typically for about 24 hours. Following the reduction, it will usually be desirable to treat the porous material to remove the organic material including the structure-directing agent, hydrocarbon additive, unreacted source material and ionic impurities, for example by solvent extraction or by decomposition in nitrogen and combustion in oxygen (calcination). However, for certain applications such treatment may not be necessary.

The regular pore structure of the porous metal may for example be cubic, lamellar, oblique, centred rectangular, body-centred orthorhombic, body-centred tetragonal, rhombohedral or hexagonal. Preferably the regular pore structure is hexagonal.

The invention is further illustrated by the following non-limiting Examples.
EXAMPLE 1 (COMPARATIVE) Preparation of 20% Porosity Mesoporous Copper-tin Powder 72 g of BC10-TX surfactant (from Nikkol) was heated until molten. To this was added a mixture containing 12.0 cm3 of 0.3 M tin(II) methanesulphonate solution (aqueous), 12.0 cm 3 of copper(II) sulphate solution (aqueous) and 0.63 g of sodium hypophosphite in 24 cm3 of deionised water. The resulting paste was stirred vigorously until homogeneous and then allowed to cool to room temperature and allowed to stand at room temperature overnight. The surfactant was removed from the resultant product via repeated washing in deionised water. The collected powder was dried in air, overnight at 60 C.

Preparation of 39% Porosity Mesoporous Copper-tin Powder 72 g of BC10-TX surfactant was heated until molten. To this was added a mixture containing 12.0 cm3 of 0.6 M tin(II) methanesulphonate solution (aqueous), 5 12.0 cm3 of 0.6 M copper(II) sulphate solution (aqueous) and 0.42 g of dimethylamine-borane complex in 24 cm3 of deionised water. The resulting paste was stirred vigorously until homogeneous and then allowed to cool to room temperature and allowed to stand at room temperature overnight. The surfactant was removed from the resultant product via repeated washing in deionised water. The collected powder was 10 dried in air, overnight at 60 C and was found to have an average pore size of 2.5 nrn.

Preparation ofHijeh Porosity Mesoporous Copper-tin - 51 % Porosity 72 g of BC10-TX surfactant was heated until molten. To this was added a mixture containing 12.0 cm3 of 1.0 M tin(II) tetrafluoroborate solution (aqueous), 12.0 cm 3 of copper(II) tetrafluoroborate solution (aqueous), 3.15 g sodium citrate, 2.23 g ethylenediaminetetraacetic acid (EDTA) and 2.11 g of sodium hypophosphite in 24 cm3 of deionised water. The resulting paste was stirred vigorously until homogeneous and then allowed to cool to room temperature and allowed to stand at room temperature overnight. The surfactant was removed from the resultant product via repeated washing in deionised water. The collected powder was dried in air, overnight at 60 C
and was found to have an average pore size of 9-10 nm.

Preparation of HiQh Porosity Mesoporous Copper-tin Powder - 44% porosity 72 g of BC10-TX surfactant was heated until molten. To this was added a mixture containing 12.0 cm3 of 1.0 M tin(II) tetrafluoroborate solution (aqueous), 12.0 cm 3 of copper(II) tetrafluoroborate solution (aqueous), 3.15 g sodium citrate, 2.23 g ethylenediaminetetraacetic acid and 2.11 g of sodium hypophosphite in 24 cm3 of deionised water. The resulting paste was stirred vigorously until homogeneous and then allowed to cool to room temperature and allowed to stand at room temperature overnight. The surfactant was removed from the resultant product via repeated washing in deionised water. The collected powder was dried in air, overnight at 60 C.
Treatment of the powder at 250 C in H2/Ar for 5 hours resulted in a powder with an average pore size of 9-10 nm.

Electrode Fabrication Using SBR/CMC Binder A lithium ion battery anode based on mesoporous copper-tin fabricated using a liquid crystal templating route was prepared. The copper-tin material had a porosity of 51 % as calculated from nitrogen porosimetry (BET) measurements and was prepared as described in Example 3. This was done by first mixing the copper-tin material with Timcal KS-6 graphite, followed by addition of an aqueous solution of styrene butadiene rubber (SBR) and carboxy methyl cellulose (CMC), such that the percentages of copper-tin, SBR, CMC and graphite in the electrode (after, evaporation of the water) were 80 %, 6 %, 4 % and 10% by mass respectively. The resulting paste was then spread over a 14 m thick copper foil which acted as a current collector, and the water was allowed to evaporate leaving a uniform coating of the copper-tin/SBR/CMC/graphite composite adhered to the copper foil. This composite electrode was then calendared to improve adhesion.

Electrode Fabrication UsinP EPDM Binder A lithium ion battery anode based on mesoporous copper-tin fabricated using a liquid crystal templating route was prepared. The copper-tin material had a porosity of 51 % as calculated from nitrogen porosimetry (BET) measurements and was prepared as described in Example 3. This was done by first mixing the copper-tin material with a solution consisting of hexane and ethylene propylene diene monomer (EPDM), followed by addition of Timcal KS-6 graphite, such that the percentages of copper-tin, EPDM and graphite in the electrode (after evaporation of the hexane) were 90 %, 5 %
and 5 % by mass respectively. The resulting paste was then spread over a 14 m thick copper foil which acted as a current collector, and the hexane was allowed to evaporate, leaving a uniform coating of the copper-tin/EPDM/graphite composite adhered to the copper foil.

Cell Fabrication Using SBR/CMC Bound Electrodes A lithium ion battery with a footprint area of 1.2 cm2 was fabricated using a home-made cell housing. The cathode consisted of LiCoO2 supported on aluminium as is standard in the industry. The anode consisted of a composite of liquid crystal templated mesoporous copper-tin, SBR/CMC and graphite deposited on a copper foil as prepared in Example 5, but using the material of Example 4. The separator consisted of two layers of Celgard 2400 membrane and contained an electrolyte composed of 1 M
LiPF6 in a mixture of ethylene carbonate and diethylene carbonate (LP30 Selectipur from Merck). A lithium foil was inserted between the two layers of separator and acted as a reference electrode. Once assembled, the cell was cycled at a C/10 rate with a depth of discharge of 30 % using a lower voltage limit of 0.005 V vs. Li/Li+
for the copper-tin composite electrode..

Similar cells were prepared, but using the materials of Example 1(Comparative) and Example 2 as the anode materials, in place of the material of Example 4.

Figure 1 compares the cycle life behaviour of the three cells; one utilising an anode as described in Example 1 (Comparative), one from Example 2 and another that used the same construction, but using the 44% porous copper-tin material prepared as described in Example 4.

Cell Fabrication Using EPDM Bound Electrode A lithium ion battery with a footprint area of 1.2 cm2 was fabricated using a home-made cell housing. The cathode consisted of LiCoO2 supported on aluminium as is standard in the industry. The anode consisted of a composite of liquid crystal templated nanoporous copper-tin, EPDM and graphite deposited on a copper foil as prepared in Example 6. The separator consisted of two layers of Celgard 2400 membrane and contained an electrolyte composed of 1 M LiPF6 in a mixture of ethylene carbonate and diethylene carbonate (LP30 Selectipur from Merck). A lithium foil was inserted between the two layers of separator and acted as a reference electrode. Once assembled, the cell was cycled at a C/10 rate with a depth of discharge of 30 % using a lower voltage limit of 2.5 V.

A similar cell was prepared, but using the material of Example 2 as the anode materials, in place of the material of Example 3.

Figure 2 compares the cycle life behaviour of the two cells; one utilising an anode as described in Example 6 and another that used the same construction, but with a 39 % porous copper-tin material prepared as described in Example 2.

Claims (14)

CLAIMS:

1. An electrochemical cell comprising a positive electrode, a negative electrode and a non-aqueous electrolyte, where the negative electrode comprises a liquid crystal templated mesoporous material capable of forming a lithium insertion alloy, characterised in that the liquid crystal templated mesoporous material has a porosity of from 38% to 80%.

2. A cell according to Claim 1, in which the porosity of the mesoporous material is from 42% to 75%.

3. A cell according to Claim 2, in which the porosity is from 44% to 70%.

4. A cell according to Claim 3, in which the porosity is from 50% to 65%.

5. A cell according to any one of the preceding Claims, in which the negative electrode comprises a mesoporous element selected from the group consisting of aluminium, silicon, magnesium, tin, bismuth, lead and antimony.

6. A cell according to Claim 5, in which said element is tin.

7. A cell according to any one of the preceding Claims, in which said negative electrode additionally comprises an element inactive for lithium insertion.

8. A cell according to Claim 7, in which said element inactive for lithium insertion is selected from the group consisting of copper, nickel, cobalt and iron.

9. A cell according to Claim 7, in which said element inactive for lithium insertion is copper.

10. A cell according to any one of Claims 7 to 9, in which said element active for lithium insertion and said element inactive for lithium insertion are alloyed.

11. A cell according to any one of Claims 7 to 10, in which said element active for lithium insertion is tin and said element inactive for lithium insertion is copper.

12. A cell according to any one of the preceding Claims, in which the liquid crystal templated mesoporous material is supported on a support.

13. A cell according to Claim 12, in which the support also functions as a current collector.

14. A cell according to Claim 13, in which the support is of copper, nickel or cobalt.