Electrochemical Cell Production
The present invention relates to a method of assembling an electrochemical cell, to a method of laminating cell components, and to electrochemical cells so made .
For many years it has been known to make cells with lithium metal anodes, and cathodes of a material into which lithium ions can be intercalated or inserted. Such cells may use, as electrolyte, a solution of a lithium salt in an organic liquid such as propylene carbonate, and a separator such as filter paper or polypropylene. The use of a gel or solid electrolyte containing a polymer, a plasticiser, and a lithium salt, has also been suggested as an electrolyte. In the case of secondary or rechargeable lithium cells, the use of lithium metal anodes is unsatisfactory as problems arise from dendrite growth, but the use of an intercalation material such as graphite has enabled satisfactory cells to be made. Such cells may be referred to as "lithium ion" cells, or "swing" cells, as lithium ions are exchanged between the two intercalation materials during charge and discharge.
Gel or solid electrolytes may be made, as described by Gozdz et al (US 5 296 318), with a copolymer of 75 to 92% vinylidene fluoride and 8 to 25% hexafluoropropylene as the polymer, this being dissolved in a low boiling- point solvent such as tetrahydrofuran along with a lithium salt and a plasticising solvent such.as ethylene carbonate/propylene carbonate mixture, and cast from solution. Such an electrolyte, but using homopolymer polyvinylidene fluoride (PVdF) with a very low melt flow index, is described in GB 2 309 703 B (AEA Technology) . GB 2 309 701 B (AEA Technology) describes how the adhesion of such an electrolyte composition can be
enhanced by grafting suitable mono-unsaturated groups onto the polymer chain, and in this case the polymeric chain might be a homopolymer PVdF, or a copolymer of vinylidene fluoride. It is also possible to make such a solid polymer electrolyte by first making a porous film of the polymer material, and then immersing the film in a solution of lithium salt in an organic solvent so the electrolyte solution is absorbed by the polymer film, as described in EP 0 730 316 A (Elf Atochem) . WO 01/48053 (AEA Technology) describes a way of making a microporous membrane of a polymer consisting at least primarily of vinylidene fluoride, and such films can be less than 50 μm thick. Such a thin electrolyte film is desirable in reducing cell resistance. However, with films less than say 30 μm thick it becomes difficult to laminate the electrolyte layer to the electrode layers without risk of shorting, when using conventional lamination techniques that require application of pressure and elevated temperature .
According to the present invention there is provided a method of making a lithium ion polymer cell comprising an anode layer and a cathode layer each comprising respective lithium ion insertion materials, separated by a porous polymeric membrane, wherein the anode layer and the cathode layer each incorporates a polymeric binder, the method comprising assembling the anode layer, the porous polymeric membrane, and the cathode layer to form a cell .assembly, wherein a thin porous layer of a copolymer material is arranged to cover the faces of the anode layer and the cathode layer, the method also comprising contacting the cell assembly with a liquid that is a solvating liquid for the copolymer material but not for the polymeric material of the membrane, and then evaporating the liquid from the cell assembly; and finally introducing an electrolyte solution into the cell
assembly to form a cell.
It is evident that the polymeric material of the membrane must be different from that of the copolymer layers, and may for example be polyethylene or polypropylene, or homopolymer PVdF, or a copolymer of a different composition and different solubility to that of the copolymer layers. The porous polymeric membrane is preferably a microporous membrane of thickness less than 50 μm, preferably less than 30 μm, for example 20 μm.
The thin copolymer layers are preferably less than 20 μm thick, more preferably less than 10 μm thick, for example 2 μm thick. Such very thin layers are not easy to handle. They may be coated onto the anode and cathode by laminating or by casting directly onto the anode or cathode, and in the latter case may be even thinner (and may also, in practice, be of non-uniirorm thickness) . They may also be formed on the anode or cathode by spraying, and may therefore be discontinous .
The liquid may be acetone. This partially solvates the surfaces of the thin copolymer films so that they adhere to the membrane. This solvation process may be assisted by keeping the cell assembly at a slightly elevated temperature such as 30 °C (and preferably not more than 100 °C) . The subsequent evaporation may be performed at a higher temperature, for example at 60°C, preferably aided by reduced pressure. For example the cell assembly may be vacuum dried at 60°C for 3 hours to ensure removal of all traces of the acetone. Instead of acetone, the liquid used may be a mixture of acetone and dimethyl carbonate, or a mixture of ethylene carbonate and dimethyl carbonate, or it may be ethyl-methyl-ketone. These effectively act as solvents for the copolymer at about 60°C.
Both the porous copolymer layers and the porous membrane are preferably microporous, with pores which are preferably between 0.1 and 10 μm across, more preferably between 0.5 and 2 μm. As described in WO 01/48063, a microporous membrane may be cast from a solvent/non- solvent mixture, or from a latent solvent, so that the entire process can be carried out in the absence of water or humidity, reducing the risk of water being present in the final film or membrane (which would be detrimental to the properties of a lithium ion cell) . The non-solvent should not only dissolve in the solvent, but it should be miscible with the solvent in substantially all proportions. The boiling point of the non-solvent is preferably higher than that of the solvent, preferably about 20 °C higher. For example the solvent might be dimethyl formamide or dimethyl acetamide, in which case a suitable non-solvent is 1-octanol which is soluble in those solvents and whose boiling point is about 194 °C.
Some liquids suitable as solvents, and as latent solvents, for vinylidene fluoride-based polymers are listed in the Tables. It should however be appreciated that not all solvents are suitable for all grades of polymer.
Table 1
The evaporation rate during drying must not be 5 rapid, as rapid drying tends to produce macropores, and also may lead to formation of an impervious skin hich prevents evaporation of underlying liquid. When using a latent solvent, the drying process should be carried out at a temperature below the dissolution temperature for
10 the latent solvent. Consequently the polymer precipitates, and it is believed that two phases occur: a polymer-rich phase, and a polymer-poor phase. As the latent solvent evaporates the proportion of the polymer- rich phase gradually increases, but the remaining
15 droplets of polymer-poor phase cause the formation of pores.
The invention will now be further described, by way of example only, and with reference to the accompanying 20 drawings in which:
Figure 1 shows graphically the variation of voltage with charge for a cell of the invention, at different rates of discharge;
25 Figure 2 shows graphically the variation of voltage with charge for another cell of the invention;
Figure 3 shows a photograph of a copolymer film coated onto an electrode; and
Figure 4 shows a photograph of an alternative copolymer film formed by spraying.
Making the porous membrane
Homopolymer PVdF (Solvay grade 1015) , which has a low value of melt flow index (about 0.7 g/10 min at 10 kg 'and 230 °C) , is dissolved in dimethyl formamide (DMF) at a temperature of 45 °C while stirring; 15 g of PVdF were dissolved in 85 g of DMF. A small quantity, 9 g, of 1- octanol is then added dropwise to the polymer solution, and carefully mixed during this addition to ensure the mixture is homogeneous. The quantity of 1-octanol must not be too large, or the solution will gel. The mixture is then mixed for a further period of 2 hours to ensure uniformity. The resulting ternary mixture is then cast, using a doctor blade over a roller, onto an aluminium foil substrate to form a layer initially 0.25 mm thick, and then passed through a 7 m long drying tunnel with two successive drying zones at temperatures of 65°C and 100°C respectively. It moves through the drying tunnel at 0.5 m/min. Within the drying zones the film is exposed to a dry air flow with a velocity of 14 m/s, to remove any solvent and non-solvent that evaporates . The dry air is obtained by passing air through a dehumidifier, so its dewpoint is -40 °C.
During passage of the film through the drying tunnel, which takes 14 minutes, both the solvent and non- solvent gradually evaporate (although they are both well below their boiling points) , the solvent tending to evaporate more rapidly. A white polymer membrane is thereby obtained, of thickness about 20 μm, and analysis
- 1 -
with a scanning electron microscope shows it to be microporous. The pores are of size in the range 0.5-2.0 μm, typically about 1 μm in diameter, at least at the surface. The membrane has been found to have a porosity of about 53%.
Making the thin copolymer layers
The thin copolymer layers are made by a similar process, a 12% by weight solution of a copolymer PVdF/6HFP (vinylidene fluoride and 6% by weight hexafluoropropylene) being made.by dissolving 12 grams of the copolymer in 88 grams DMF. A small quantity of 1- octanol is then added dropwise to the copolymer solution, and carefully mixed during this addition to ensure the mixture is homogeneous, and is then kept stirred for 2 hours. The resulting ternary mixture is then cast, using a doctor blade over a roller, onto an aluminium foil substrate to form a layer initially 0.06 mm thick, and then dried exactly as described above.
This forms a microporous layer of thickness about 2 μm, the pores being similar to those in the membrane described above.
Making the electrodes
A cathode is made by making a mixture of spinel LiM 2θ4, a small proportion of conductive carbon, and homopolymer PVdF 1015 as binder (as described above) this being cast from solution in N-methyl pyrrolidone (NMP) which is a solvent for the PVdF. The mixture is cast using a doctor blade onto an aluminium foil, and passed through a dryer with temperature zones at for example 80 "C and 120 °C to ensure evaporation of the NMP. This
process is then repeated to produce a double-sided cathode. Removal of all the NMP is further ensured by subsequent vacuum drying.
An anode is made from a mixture of mesocarbon microbeads of particle size 10 μm, heat treated at 2800°C (MCMB 1028) , with a small amount of graphite, and homopolymer PVdF 1015 as binder. This mixture is cast from solution in NMP, onto a copper foil, in a similar fashion to that described in relation to the cathode.
Cell Assembly
The cathode is sandwiched between two of the thin copolymer layers so that each surface is completely covered, and these components are laminated together by subjecting them, when placed between release papers, to compressive pressure in a press between rollers giving a co pressive force of 20 N at an elevated temperature of 120°C.
The anode is also sandwiched between two thin copolymer layers, and laminated together in the same way.
A cell assembly is then wound with the porous membrane of thickness 20 μm separating the anode from the cathode. Each such cell assembly is enclosed in a sealed envelope of aluminium/plastic laminate, and a small quantity such as 0.5 g of acetone is injected into the envelope. The envelope containing the cell assembly is then held at a temperature of 30°C for a period of at least 5 minutes. This elevated temperature enhances the solvation of the surface of the copolymer layer by the acetone.
After cooling to ambient temperature, the cell
assembly is removed from the envelope and is then vacuum dried at 60 °C for 3 hours to ensure removal of any traces of acetone.
The cell assembly is then vacuum filled with a plasticising liquid electrolyte, for example 1 molar
LiPFβ in an ethylene carbonate/ethyl methyl carbonate mixture. After storing for 16 hours to ensure the electrolyte has been absorbed by all the cell components, it is then vacuum packed in a flexible packaging material.
It has been found that the anode and cathode are both laminated to the porous membrane. It is apparent that this is because the copolymer layers, when partially solvated by the acetone at 30 °C, are sufficiently tacky that they adhere to the porous membra e. Because the lamination occurs without application of external pressure there is no risk of perforation of the porous membrane. Surprisingly the partial solvation does not affect the porosity of the copolymer layers, and the overall process does not affect the porosity of the membrane, so that the cell has good electrical properties after addition of the plasticising liquid electrolyte.
For example, referring to Figure 1, this shows the variation of voltage with capacity for a cell of the invention, for various different rates of discharge, the cell being charged and discharged between voltages of 2.75 V and 4.25 V. The rated cell capacity is the capacity at a discharge current (in amps) numerically equal to a fifth of the cell capacity (in amp hours) ; this discharge is referred to as discharge at the C/5 rate. The rated cell capacity was determined from the first five discharge and recharge cycles, which were
carried out at an estimate of the C/5 rate. The measurements shown in Figure 1 were then obtained. It will be observed that the discharge capacity is about 0.62 Ah for this particular cell; the capacity decreases slightly as the rate of discharge increases; but that even at a discharge rate of 2C the available capacity is about 0.52 Ah, which is about 84% of the rated capacity.
Alternative cell
An anode and cathode are made by the same process as described above, although in this case the active insertion material in the cathode is LiCoθ2 (from FMC
Corp.) As before, both the electrodes are double-sided. The electrodes are then sandwiched between two of the thin microporous copolymer layers (as described above) and laminated between rollers at an ej-evated temperature as in the previous example.
A cell assembly is then wound with a microporous polyethylene membrane of thickness 16 μm separating the anode from the cathode, this membrane being supplied by Tonen Chemical Corp. Each such cell assembly is enclosed in a sealed envelope, and a small quantity such as 0.5 g of acetone is injected into the envelope. The envelope containing the cell assembly is then held at a temperature of 30°C for a period of at least 5 minutes. This elevated temperature enhances the solvation of the surface of the copolymer layer by the acetone.
After cooling to ambient temperature, the cell assembly is removed from the envelope and is then vacuum dried at 60 °C for 3 hours to ensure removal of any traces of acetone.
The cell assembly is then vacuum filled with a plasticising liquid electrolyte consisting of 1 molar
LiBF in an solvent mixture comprising 60.83 (weight) % gamma-butyrolactone, 24.33% ethylene carbonate, 12.16% methoxyethyl methyl carbonate and 2.68% vinyl ethylene carbonate. After storing for 16 hours to ensure the electrolyte has been absorbed by all the cell components, it is then vacuum packed in a flexible packaging material. As with the previous cell, the anode and cathode are both laminated to the porous membrane in this case .
Cells made in this way were charged, and aged for two weeks, before their capacity was measured as described above. Referring to figure 2, this shows the variation of voltage with capacity for a laminated cell, for various different rates of discharge, the cell being charged and discharged between voltages of 2.75V and 4.25V. The rated cell capacity in this case was about 0.66 Ah for this particular cell. As with the cell described above, the capacity decreases slightly as the rate of discharge increases; but even at a discharge rate of 2C the available capacity is about 95 percent of the rated capacity.
It will be appreciated that the cell lamination may use a different liquid. For example, instead of acetone, the liquid might be a 50:50 (by weight) mixture of acetone and dimethyl carbonate, the latter component not acting as a solvent itself, and acting as a diluent for the solvent. The cell assembly, consisting of the electrodes and the microporous membrane wound together, may be briefly dipped into this mixed liquid, removed, and immediately subjected to vacuum drying at 60 °C. The liquid wets all the surfaces of the cell components, and
lamination takes place as with the previously-described procedure. This procedure avoids the need for the use of the sealed envelope, and has been found to give equally good results.
Another suitable- liquid would be a mixture of ethylene carbonate and dimethyl carbonate, which acts as a latent solvent for the copolymer at about 60 °C. The cell assembly may be dipped into this mixed liquid, removed, and immediately subjected to vacuum drying at 60 °C to laminate the cell.
Rather than forming the microporous copolymer layers as separate items, they may instead be formed in situ, for example being cast directly onto the surfaces of the anode or the cathode. In this case the layer may be even thinner, as it doesn't have to be handled. The layer might for example be only 1 μm thick, and may be somewhat non-uniform in its thickness; indeed, as a result of non- uniformities in the surface of the anode or cathode there may be macroscopic pores in the copolymer layer. Referring to figure 3, there is shown (to approximately full scale) a roll of electrode material onto which a thin copolymer film has been cast. The width of the material is 10 cm, and it will be noted that the bulk of the film is white, indicating that it is microporous, but that there are darker spots where the film is thinner or non-existent.
Alternatively the microporous copolymer layer may be formed by spraying onto the electrode surfaces. Referring to figure 4 there is shown (to approximately full-scale) a roll of electrode material onto which copolymer material has been sprayed, forming a discontinuous layer. The white spots indicate that at least those regions of the sprayed film are microporous.
Cells made as described earlier but using copolymer layers formed directly on the surfaces of the electrodes, as shown in figures 3 and 4, have been found to have electrical properties substantially the same as those described earlier in relation to figure 1 and figure 2.