Non-woven fabrics can be made by processes which include (a) melt blowing, (b) spinning, and (c) wet or dry laying.

The fibres of fabrics made by spinning and wet or dry laying require bonding to one another for the fabric to have integrity, so that it has the mechanical properties required for satisfactory performance. In the case of fabrics made by spinning, the fibres are bonded to one another by the application of heat and pressure. In the case of fabrics made by wet or dry laying, polyethylene is incorporated into the fabric, either as fibres consisting essentially of polyethylene or as bicomponent fibres consisting of a polypropylene core and a polyethylene sheath. The polyethylene in the fabric can provide the necessary bonds as a result of heating the fabric to a temperature that is greater than the softening point of the polyethylene.

Non-woven fabrics can be used to form an electrode separator in an electrochemical device. Examples of such devices include nickel-cadmium and nickel-metal hydride cells. The separator should be inert towards materials with which it comes into contact in the cell including in particular the alkaline electrolyte and the electrode materials. It should also have physical characteristics which enable it to withstand the treatment encountered during assembly of the device and during use. For example, it should be able to withstand the stresses encountered during spiral winding of the cell components. It should also be capable of resisting the growth of dendrites between the electrodes during recharging.

A fabric that is made from spun fibres which are then bonded together (a "spun bonded" fabric) has the disadvantage that the bonds reduce the effective surface area of the fabric that is available to ion transfer by effectively blocking the pores of the fabric. The uneven current distribution that results from this uneven pore distribution can give rise to dendrite formation during recharging of a secondary cell, ultimately leading to a short circuit in the cell. There is therefore a compromise to be reached with such fabrics between mechanical properties that are enhanced by bonds between the fibres and aspects of electrochemical performance which can be diminished by the bonds.

Furthermore, manipulation of a fabric formed from spun fibres (whether or not bonded to one another), in particular prior to and during assembly of a device with the fabric as a separator, can lead to deformation of the fabric involving relative movement of the fibres of the fabric (in regions between bonds when present). This movement can involve for example untangling of the fibres. It results in opening of the structure of the fabric. This has the significant disadvantage of reducing the effectiveness of the fabric as a barrier and increasing the risk of failure of a device in which the fabric is incorporated as an electrode separator, generally by shorting of the electrodes.

The structure of a melt-blown fabric is stable when placed under stress. A melt-blown fabric has the further advantage of small fibres size (which is generally less than about 5 pm, and often as low as 1 pm or less) so that the fabric can provide a separator with good barrier properties.

However, the fine size of the fibres of the fabric means that the fabric is only able to withstand the application of small stresses and small degrees of strain. A large stress or high strain can result in fracture of the fabric.

Non-woven fabrics formed by wet or dry laying of fibres can have satisfactory mechanical properties. However, especially when bicomponent fibres are used, the fibre size can tend to be undesirably large, often greater than 15 um. A further disadvantage which arises from the use of bicomponent fibres is their high cost.

The present invention provides a laminate of melt-blown and spun fibre fabrics, the fibres of which have been treated by graft copolymerisation of a vinyl monomer such as acrylic acid.

Accordingly, in one aspect, the invention provides a laminate formed from first and second non-woven fabrics which each comprise fibres of a hydrophobic polymeric material, the first fabric being formed from spun fibres and the second fabric being a melt-blown fabric, the fibres of the fabrics having undergone a copolymerisation reaction with a vinyl monomer which is capable of reacting with an acid or a base to form a salt directly or indirectly, the reaction involving exposure of the laminate to ultraviolet radiation while impregnated with a solution of the vinyl monomer resulting in grafting of. the vinyl monomer to the surfaces of the fibres.

The laminate of the invention has the advantage that it is able to tolerate stresses imposed when it is being manipul- ated, for example prior to and during assembly of an electro- chemical device in which the laminate is incorporated as an electrode separator. In particular, the structure of the laminate remains stable under moderate loads and does not exhibit the tendency to open as in the case of fabrics formed from spun fibres. Furthermore, the laminate structure has a reduced tendency to fracture when placed under load compared with melt-blown fabrics.

A further advantage of the laminate of the invention is that it can exhibit the good barrier properties which can be obtained from melt-blown non-woven fabrics when it is used as an electrode separator in an electrochemical device. This arises from the small effective pore size that is presented by the laminate. The effective size of the pores defined by the fibres of the fabrics can be measured using a Coulter porometer. Preferably, the effective pore size of the laminate is less than about 30 pm, more preferably less than about 20 pm, for example less than about 15 pm. A small pore size has the advantage of enhancing the ability of the laminate to resist penetration of electrode materials, for example as dendrites. The laminate of the invention can be subjected to a calendering step during its manufacture.

Amongst other advantages, this can have the result of reducing the effective pore size of the laminate.

A small pore size also enhances the ability of the laminate to absorb and retain electrolyte once the fibres have been treated to render them hydrophilic. A high electrolyte absorption has the advantage of reducing the internal resistance of an electrochemical device in which the laminate is incorporated as an electrode separator, and of extending the cycle life of the device.

Yet another advantage of the laminate of the present invention arises from the fine structure presented by the melt-blown fabric when the laminate is used as an electrode separator. The laminate is able to combine the good physical properties discussed above with an ability to absorb contaminants. These contaminants, including ammonia and metal ions, can be found in electrochemical devices following the production and use of certain electrode materials, for example nickel hydroxide and metal hydride electrodes.

Absorption of contaminants in the device has the advantage of inhibiting self-discharge reactions. The shelf-life of a device with a laminate of the present invention as its separator can therefore be enhanced compared with devices with previously known separators.

In the case of ammonia which can be formed by the reduction of nitrate ion contamination from a nickel hydroxide electr- ode, it has been found that there is a residual ion exchange capacity in addition to the ion exchange capacity that is measured using standard techniques such as the titration of the acid form of the membrane with potassium hydroxide. It is believed to be this residual ion exchange capacity that enables ammonia and other contaminants to be absorbed in an electrochemical device. The residual ion exchange capacity is expressed in terms of the milliequivalents per gram of the laminate, and is measured as described below. Preferably, it is at least about 0.15 meq.g-l, preferably at least about 0.25 meq.g-l, more preferably at least about 0.3 meq.g-l, for example at least about 0.35 meq.g-l.

The laminate of the present invention also has the advantage of reduced cost compared with products based on bicomponent fibres such as bicomponent polyethylene and polypropylene fibres.

The fibres of the first fabric can be bonded to one another prior to deformation of the laminate, for example by localised welds between the fabrics. The first fabric might then be a spun bonded fabric.

The fibres of the first fabric can be substantially unbonded to one another in which the fabric is formed without a step of bond formation by the application of heat and pressure.

There might be weak forces between the fibres of such fabrics. For example, weak forces can result from a step of calendering a fabric under moderate heat and pressure, which can lead to localised deformation of the fibre material, especially where fibres come into contact with one another.

However, the forces will be capable of being overcome when the fabric is placed under tension. It will be possible to discern a boundary between the fibres of the fabric. There will not be any intimate mixing of the materials of the fibres as results from the formation of a weld. Features of treated fabrics formed from unbonded fibres are disclosed in the patent application filed with the present application, claiming priority from UK patent application no. 9712690.8 and entitled NON-WOVEN FABRIC TREATMENT (bearing the agents' reference P10599). Subject matter disclosed in the specification of that application is incorporated in the present specification by this reference.

Preferably, there are bonds between the first and second fabrics of the laminate. For example, the bonds might be formed by localised application of heat and pressure. The heat and pressure can be applied by passing the laminate between heated rollers with appropriately profiled surfaces.

Such treatment can lead to the formation of localised welds between the fabrics. They will also tend to form bonds between the fibres of the first fabric (which might already be bonded to one another prior to lamination). Preferably, the proportion of the area of the laminate in which the bonds are formed is less than about 20%, more preferably less than about 15%, especially less than about 10%, for example about 8%.

Preferably, the mean thickness of the fibres of the first fabric (which might be measured as a mean diameter, especially when the fibres have a circular cross-section) from which the non-woven fabric is formed is not more than about 30 pm, more preferably not more than about 20 pm. The thickness of the fibres of the first fabric will often be at least about 5 pm, for example at least about 10 pm.

Preferably, the mean thickness of the fibres of the second fabric (which might be measured as a mean diameter, especially when the fibres have a circular cross-section) from which the non-woven fabric is formed is not more than about 8 pm, more preferably not more than about 5 pm. The thickness of the fibres of the second fabric will generally be at least about 0.5 pm.

Preferably, the ratio of the weight of the fibres of the second fabric to the weight of the fibres of the entire laminate is at least about 0.1, more preferably at least about 0.2, especially at least about 0.4, for example at least about 0.5.

Use of fabrics with different constructions or different fibres or both can lead to a laminate with asymmetric properties. This can lead to differing extents of the copolymerisation reaction from one side of the laminate to the other. The presence of a surface of the laminate which is relatively less hydrophobic than another surface has advantages when the laminate is for use as a separator in certain types of electrochemical devices. The hydrophilic surface is more easily wetted by aqueous electrolyte which can be advantageous in the region of a positive electrode.

This can inhibit flow of hydrogen gas from the negative electrode to the positive electrode when the battery is being recharged. The less hydrophilic surface lead to the creation of a three phase boundary at the surface of the negative electrode (between the surface of the electrode, the electro- lyte and oxygen gas generated at the electrode), which can help to control the internal pressure in the battery on recharging due to the generation of oxygen.

The laminate of the invention can include one or more fabrics in addition to the first and second fabrics. For example, the laminate of the invention can include the first and second fabrics as discussed above, together with a third fabric and possibly a fourth fabric. The third fabric can be formed from spun fibres. The third fabric can have the same construction as the first fabric. Spun fibres of a third fabric can be bonded to one another prior to formation of the laminate, that is as a spun bonded fabric. Generally, however, the fibres of a third fabric will be bonded to one another by localised application of heat and pressure by which bonds between the fabrics of the laminate are formed.

The first fabric and a third fabric formed from spun fibres can be arranged on opposite faces of the second fabric.

The ion exchange capacity of the laminate is measured in meq.g-l according to the test routine referred to below, to provide a measure of the extent of the graft copolymerisation reaction between of the material of the fibres and the vinyl monomer. Preferably, the ion exchange capacity is at least about 0.25, more preferably at least about 0.4, especially at least about 0.6. Preferably, the ion exchange capacity is not more than about 2.0, more preferably not more than about 1.6, especially not more than about 1.4, for example not more than about 1.2. It has been found that useful increases in the physical properties of a non-woven fabric, in particular when formed from polypropylene fibres, can be obtained at low graft levels corresponUing to these values of the ion exchange capacity.

The gel fraction of the material of the laminate is measured according to ASTM D2765-84, providing a measure of the extent of crosslinking of the material of the fibres. Preferably, the gel fraction is at least about 10%, more preferably at least about 20%, especially at least about 30%.

Preferably, the thickness of the laminate, measured using test method DIN 53105 which involves lowering a 2.0 kg weight onto a sample of the laminate of area 2.0 cm2 at a speed of 2.0 mm.s-l, is greater than about 80 pm, more preferably greater than about 100 pm; preferably, the thickness is less than about 400 pm, more preferably less than about 250 pm.

The method by which the laminate is made may include a calendering step to reduce its thickness to a value within the range referred to above, the reduction being by at least about 5%, preferably at least about 15%, more preferably at least about 25%, and less than about 60%, preferably less than about 45%, more preferably less than about 40%. Calend- ering can have the advantage of reducing the effective size of the pores in the fabric, giving rise to the advantages discussed above. The calendering step may take place before or after the material of the laminate is reacted with the graft-polymerisation solution. Calendering the laminate before the graft-polymerisation reaction has been found to give rise to increased rates of the reaction. Calendering the laminate after the graft-polymerisation reaction has been found to give rise to enhanced electrolyte absorption. A laminate that has been calendered after the graft reaction can have an improved ability to absorb impurities, especially ammonia, which might be present in the electrolyte system.

Moreover, fibres of the laminate are less likely to be damaged physically as a result of the calendering step when it is carried out after the graft-polymerisation reaction.

The vinyl monomer which is graft-polymerised with the material of the fibre surface can be capable of reacting with an acid or a base directly to form a salt, or indirectly to form a salt after appropriate work up, perhaps involving for example hydrolysis or sulphonation. Preferred vinyl monomers include ethylenically unsaturated carboxylic acids and esters thereof such as acrylic acid, methacrylic acid, methyl acrylate, and methylmethacrylate. Other vinyl monomers which might be used include acrylamide, vinylpyridine, vinyl- pyrrolidone and styrene-sulphonic acid.

Preferably, the material of the surface of at least some of the fibres, for example at least about 40% by weight, pref- erably at least about 60%, more preferably at least about 80%, comprises polypropylene. Preferably, at least 40% by weight of the material of the fibres of the first fabric or the second fabric or both is polypropylene, more preferably at least about 60%, especially at least about 80%.

Preferably, the material of at least some of the fibres from which the first or second fabric (or each of the fabrics) is formed, for example at least about 40% by weight, preferably at least about 60%, more preferably at least about 80%, is substantially homogeneous throughout the thickness of the fibres. It can be preferred for many applications for the material of substantially all of the fibres to be substan- tially homogeneous throughout their thickness, so that those fibres are formed only from polypropylene or another suitable material (with appropriate additives where necessary).

In another aspect, the invention provides a method of treating a laminate formed from first and second non-woven fabrics which each comprise polymeric fibres, in which the first fabric is formed from spun fibres and the second fabric is a melt-blown fabric, the method comprising: (a) impregnating the laminate with a solution of a vinyl monomer which is capable of reacting with an acid or a base to form a salt directly or indirectly, the solvent being one which does not evaporate signif- icantly in the subsequent step of exposing the fabric to radiation, and (b) exposing the impregnated laminate to ultraviolet radiation while the exposure of the laminate to oxygen is restricted, to cause the monomer and the material of the fibres to co-polymerise.

The ultraviolet radiation initiated polymerisation reaction can be completed surprisingly quickly, for example by exposing the impregnated laminate to radiation for as little as 15 seconds, even as little as 5 or 10 seconds, and it has been found that the fabrics of the laminate after reaction contain a significant amount of grafted monomer, which can be sufficient for the fabrics to be rendered wettable by aqueous solutions such as might be found in certain electrochemical devices. This is to be contrasted with techniques in which graft-copolymerisation reactions are initiated using, for example, electron bombardment (either of impregnated fabric or of fabric prior to exposure to monomer solution), where reaction times of many minutes can be required in order to obtain a significant degree of grafting, and even after reaction times of this order, the degree of grafting reaction can be too low for many applications. Such techniques can lead to homopolymerisation of the vinyl monomer and degrad- ation of the material of the fabric or of the vinyl monomer or both. They do not therefore lend themselves to continuous processing in the manner of the present invention.

Details of the process«for rendering the polymeric fibres of a non-woven fabric hydrophilic, involving impregnation with a solution of the vinyl monomer followed by radiation, are disclosed in WO-A-93/01622. Subject matter disclosed in that document is incorporated in the specification of this applic- ation by this reference.

The exposure of the impregnated laminate to oxygen is restricted during the irradiation, for example, by carrying out the ultraviolet irradiation step in an inert atmosphere such as an atmosphere of argon or nitrogen, or by sealing the impregnated laminate between sheets of material which are impervious to oxygen, but are transparent to ultraviolet radiation of appropriate wavelength for initiating the co- polymerisation reaction.

Preferably, the impregnation solution includes an initiator for the polymerisation reaction. Preferably, the initiator initiates the reaction by abstracting an atomic species from one of the reacting materials, for example by abstracting a hydrogen atom from polypropylene of the fabric fibres to create a polymeric radical. Following such abstraction, the polymeric radical, in contact with the monomer in solution, can initiate the formation of a grafted branch. When an atom is abstracted from the polymer of the fabric fibres, the activated polymer can react either with another polymer molecule so that the material of the fabric becomes cross- linked, or with the vinyl monomer in a co-polymerisation reaction. An example of a suitable initiator is benzo- phenone. The mole ratio of the vinyl monomer to the initiator is preferably at least about 50, more preferably at least about 100, especially at least about 175; the ratio is preferably less than about 1500, more preferably less than about 1000, especially less than about 500, more especially less than about 350; for example the ratio may be about 200.

The impregnation solution may include a component by which homopolymerisation of the vinyl monomer is inhibited.

Examples of suitable inhibitors include iron (II) and copper (II) salts which are soluble in the reaction medium, a preferred material for aqueous media being iron (II) sulphate. It has been found, however, that the need for an inhibitor can be avoided by selection of an appropriate solvent for the graft polymerisation reaction which can restrict the speed and degree of the homopolymerisation reaction, for example as a result of its ability to act as a heat sink. This can be an advantage when it is desired to minimise the amount of contaminants in the laminate.

The impregnation solution may include additional components to optimise reaction conditions such as surfactants to ensure that the solution fully impregnates the laminate, an approp- riate mixture of solvents to ensure homogeneity of the solution, and so on.

A benefit of the present invention is that physical proper- ties of the treated laminate (in particular, its tensile strength or its ability to be wetted by aqueous solutions or both) can be stable on prolonged exposure to an alkaline solution. A laminate with stable physical properties is particularly appropriate for use as a separator in electro- chemical devices in which the electrolyte comprises an alkaline solution. A test to determine stability on exposure to alkaline solution involves storing a sample of a laminate to a solution containing 30% by weight of potassium hydroxide at 71"C for 21 days, and then comparing the selected property of the exposed laminate to that of a fabric that has not been exposed to the alkaline solution.

In a further aspect, the invention provides an electro- chemical device, comprising an anode, a cathode, a quantity of an electrolyte, and an electrode separator of the type discussed above. Preferably, the cathode in the device comprises nickel (II) hydroxide. An example of material which can form the anode in such a device includes cadmium.

Alternatively, the anode may be a metal hydride electrode.

Other types of electrochemical device in which the separator of the invention finds application include secondary cells such as lead-acid cells.

Measurement of ion exchange capacitv A sample of a non-woven fabric weighing about 0.5 g is converted into the acid (H+) form by immersion in 1.0 M hydrochloric acid at 600C for 2 hours. The sample is washed in distilled water until the washing water shows a pH in the range of about 6 to 7. The sample is then dried to constant weight at 700C.

The dried sample is placed in a 100 ml polyethylene bottle to which is added accurately 10 ml of approximately 0.1 M potassium hydroxide. Additional distilled water can be added to immerse the sample fully. A further 10 ml of potassium hydroxide is added to a second polyethylene bottle, together with the same amount of distilled water as that added to the bottle containing the sample. Both bottles are stored at 600C for at least two hours.

After being allowed to cool, the contents of each bottle are transferred to glass conical flasks, and the amount of potassium hydroxide in each is determined by titration with standardised 0.1 M hydrochloric acid, using a phenolphthalein indicator.

The ion exchange capacity, measured in milliequivalents per gram, of the membrane in the dry acid (H+) form is calculated according to the equation: t2 - t1 IEC = 10W where t1 is the titration value of HCl from bottle with the sample, t2 is the titration value of HCl from bottle without the sample, and W is the weight of the dried membrane in acid (H+) form.

Measurement of residual absorption capacity A weighed sample of a non-woven fabric laminate is converted into the potassium salt form by immersion in 0.1 M KOH for about 1 hour at 70°C. The sample is washed in distilled water to remove excess KOH. Excess water is removed using a paper towel.

An ammonia solution is prepared by mixing 120 ml of distilled water and 5 ml 0.3 M NH3. The sample is immersed in the solution and placed in an oven at 400C for 2 hours. The sample is then allowed to cool.

A 100 ml sample of the solution in the flask is then titrated with 0.1 M HCl to neutrality using methyl red as an indicator.

A control reading is obtained from the solution of ammonia in distilled water, without the laminate sample.

Examples of the manufacture of an electrode separator from a non-woven polypropylene fabric is set out below.

COMPARATIVE EXAMPLE 1 A spun bonded non-woven polypropylene fabric with a thickness of 316 um, a fibre size of about 15 to 20 um, and a basis weight of 60 g.m-2, was densified by passage through a set of smooth rollers which were heated to a temperature of 125"C.

Its thickness following densification was 170 um.

The fabric was impregnated with a solution of acrylic acid by passing the fabric around rollers located in a chamber with an atmosphere of nitrogen so that the fabric passed through the solution. The solution was formulated as follows (percentages by weight) Component wt. Acrylic acid 30.0 Benzophenone 0.25 Surfactant (Lutensol ON70T) 0.5 Water 69.25 The impregnated fabric was maintained in an atmosphere of nitrogen and passed through an irradiation chamber defined by quartz glass walls. Medium pressure mercury vapour lamps were positioned parallel to one another on opposite sides of the chamber outside the quartz glass walls. The lamps had a power output of 120 W.cm1 and were located 16 cm from the fabric. Each lamp provided a parallel ultraviolet light beam with a width of 10 cm. The total exposure time of the fabric to the radiation was about 6 seconds.

The fabric was then washed in de-ionised water to remove unreacted components and then dried in an air oven at approximately 70"C.

The properties of the treated fabric are set out below, and compared with the corresponding properties of the polypropylene fabric starting material: Unsrafted Grafted Ion exchange capacity (meq.g-1) 0 0.72 Gel content (%) (ASTM D2765-84) 0 53.3 Machine direction tensile strength (N.m-1) (ASTM D882) 2800 3100 Machine direction elongation (%) (ASTM D882) >70 45 Electrolyte wicking rate (time) 60s 600s 60s 600s (30% w/w KOH) (DIN 53924-78) (mm) 0 0 45 80 Electrolyte absorption (%) (AD 447301 US Air Force Manual) Non-wetting 290 Residual ion exchange capacity (meq.g ) 0 0.28 COMPARATIVE EXAMPLE 2 A melt blown non-woven polypropylene fabric with a thickness of 200 um, a fibre size of about 3 to 5 um, and a basis weight of 46 g.m2 was impregnated with an acrylic acid solution and irradiated as described above in Comparative Example 1.

The properties of the treated fabric are set out below, and compared with the corresponding properties of the polypropylene fabric starting material: Ungrafted Grafted Ion exchange capacity (meq.g-l) 0 0.75 Gel content (%) (ASTM D2765-84) 0 43 Machine direction tensile strength (N.m1) (ASTM D882) 0.8 Machine direction elongation (%) (ASTM D882) 11 8 Electrolyte wicking rate (time) 60s 600s 60s 600s (30% w/w KOH) (DIN 53924-78) (mm) 0 0 39 132 Electrolyte absorption (%) (AD 447301 US Air Force Manual) Non-wettinga 550 Residual ion exchange capacity (meq. g-l) 0 0.48 EXAMPLE 1 A laminate was formed from a melt blown polypropylene fabric and a fabric formed from spun fibres. The melt blown fabric had a basis weight of 14 g.m2 and a fibre size of about 3 to 5 pm. The spun fibre fabric had a basis weight of 36 g.m-2 and a fibre size of about 15 to 20 pm. The laminate was created by the localised application of heat and pressure to form localised welds between the fibres of the fabrics. The laminate had a thickness of 294 um. It was densified by passage through a set of smooth rollers which were heated to a temperature of 125"C. Its thickness following densification was 170 um.

The laminate was impregnated with an acrylic acid solution and irradiated as described above in Example 1.

The properties of the treated laminate are set out below, and compared with the corresponding properties of the polypropylene laminate starting material: Grafted Grafted Ion exchange capacity (meq.g1) 0 0.4 Machine direction tensile strength (N.m-l) (ASTM D882) 1770 2400 Machine direction elongation (%) (ASTM D882) 41 19 Electrolyte wicking rate (time) 60s 600s 60s 600s (30% w/w KOH) (DIN 53924-78) (mm) 0 0 50 110 Electrolyte absorption (%) (AD 447301 US Air Force Manual) Non-wetting 240 Residual ion exchange capacity (mes.s-l) 0 0.4 EXAMPLE 2 A laminate was formedbfrom a melt blown polypropylene fabric, and two fabrics formed from spun fibres arranged on opposite surfaces of the melt blown fabric. The melt blown fabric had a basis weight of 12 g.m2 and a fibre size of about 3 to 5 pm. The spun fibre fabric had a basis weight of 19 g.m-2 and a fibre size of about 15 to 20 pm. The laminate was created by the localised application of heat and pressure to form localised welds between the fibres of the fabrics. The laminate had a thickness of 229 pm.

The laminate was impregnated with an acrylic acid solution and irradiated as described above in Example 1.

After the irradiation and washing steps, the laminate was passed through smooth rollers heated to 970C to reduce its thickness further, to 151 pm.

The properties of the treated laminate are set out below, and compared with the corresponding properties of the polypropylene laminate starting material: Unqrafted Grafted Ion exchange capacity (meq.g-l) 0 0.75 Gel content (%) (ASTM D2765-84) 0 52.7 Machine direction tensile strength (N.m-l) (ASTM D882) 2070 2660 Machine direction elongation (%) (ASTM D882) 74 55 Electrolyte wicking rate (time) 60s 600s 60s 600s (30% w/w KOH) (DIN 53924-78) (mm) 0 0 22 70 Electrolyte absorption (O) (AD 447301 US Air Force Manual) Non-wetting 335 Residual ion exchange capacity (meq. g-1) 0 0.37 USE IN A BATTERY An AA size alkaline spirally wound nickel-metal hydride (Misch metal electrode) cell was dismantled and its electrodes reduced in area by approximately 30%. Its separator was replaced by one made in accordance with Example 2 above. The cell was reassembled with 30% KOH electrolyte.

The cell was repeatedly charged at 350 mA and discharged through a 10 ohm passive load. The cell was found to be capable of delivering 600 mA.h to a 1.0 V cut-off on discharge.

STRAIN RATE AND ELONGATION TO BREAK Comparisons were made between the deformation characteristics of the fabrics of Comparative Examples 1 and 2 and the laminate of Example 2 above. The experiments were carried out using a Lloyd Instruments tensile tester, model LRX 10030. The elongation (%) and force (N) to break were recorded using different strain rates. Samples were taken across the width of the fabrics and measured 50 mm wide and 200 mm long. Strain COMP EXAMPLE 1 COMP EXAMPLE 2 EXAMPLE 2 rate (mm.min1) Extn Force Extn Force Extn ~ Force 50 0 58 183 22 52 60 153 250 56 194" 18 54 53 162" 500 52 205 16 59 44 154 1000 46 186 10 68 36 140 (a - force to yield) These results indicate that, at high strain rates such as when assembling battery components by winding, a melt blown fabric is unsuitable because of the poor ability to withstand tensile forces. The spun fibre fabric and the laminate have a greater ability to withstand tensile forces.

STRAIN RATE AND ELONGATION The materials used in the experiment to assess the effect of strain rate to a pre-determined extension on tensile performance were analysed to determine the effect on the maximum pore size of the fabrics. Pore sizes were measured in pm using a Coulter porometer. Strain Elongn (%) Maximum pore size (pm) rate (mm.min-l) COMP EX 1 COMP EX 2 EX 2 0 0 76.72 20.04 14.43 1000 5 90.14 20.85 15.29 1000 7 90.14 2085 20.85 1000 10 109.2 fail 19.73 These results indicate that the melt blown fabric of Comparative Example 2 can maintain its structure at high strain rates but only to limited extension. It is therefore poorly suited to incorporation in an electrochemical device by winding.

The spun fibre fabric of Comparative Example 1 can withstand large deformation at high strain rates. However, the pore structure is disrupted as a result.

The pore structure of the laminated structure of Example 2 is maintained at high deformation and at high strain rates.