Nonwovens as Separators for Alkaline Batteries

Journal of The Electrochemical Society, 154 �5� A481-A494 �2007� A481

Downl

Nonwovens as Separators for Alkaline BatteriesAn OverviewPeter Kritzera,z and John Anthony Cookb,*aFreudenberg Dichtungs- und Schwingungstechnik GmbH & Co KG, Innovation Center, D-69465Weinheim, GermanybFreudenberg Nonwovens LP, Swindon SN3 5HY, United Kingdom

An overview of the manufacturing processes and performance requirements of nonwoven separators used in primary and second-ary alkaline battery systems is presented. The systems described are alkaline manganese �Alk-Mn�, nickel-cadmium �NiCd�,nickel-metal-hydride �NiMH�, nickel-zinc �NiZn�, and zinc-air batteries. The separators used in these systems consist totally orpartially of nonwoven materials. Accordingly, the paper starts with an overview of the typical methods of nonwoven production�web formation, bonding, and post-treatment� and how the products behave within the different battery systems. The paper thendiscusses the advantages and disadvantages of the different nonwoven products as battery separators in the various electrochemicalsystems.© 2007 The Electrochemical Society. �DOI: 10.1149/1.2711064� All rights reserved.

Manuscript submitted August 16, 2006; revised manuscript received December 29, 2006. Available electronically April 2, 2007.

0013-4651/2007/154�5�/A481/14/$20.00 © The Electrochemical Society

While a number of review articles dealing with separators usedin lead acid and lithium batteries have been published within the lastfew years,1-4 there have been no papers covering separators exclu-sively for alkaline systems.

The aim of the present paper is to fill this gap. It is not theprimary purpose of this paper to show the newest separator trendsfor some niche applications, but to give a general overview for theseparator as a battery raw material, both from the separator applica-tion and separator production side.

Alkaline battery systems have been under-represented in the cur-rent scientific literature, but their market potential is high. In theU.S. alone, the market for alkaline batteries, both primary and sec-ondary, was �$5.5 billion in 2005. For example, the NiMH second-ary system is being used in a number of new applications such ashybrid electric vehicles �HEVs� and as a replacement for NiCd cellsin certain applications, because of the potential ban of the cadmiumbased technologies. It can also be assumed that the current nichesystems like NiZn or Zn-air will increase their presence, if the stillcurrent problems are overcome.

Although a separator is a “dead” material inside the battery, andit only reduces the specific energy of the cell, it is not just an “openfoil”. On the one hand, the separator has direct influence on the keyparameters of the battery such as charge efficiency and dischargecapacity, and on the other hand, the wrong separator will lead topremature failure of the battery. Additionally, some of the basicrequirements for a separator are contradictory. While from a chargeacceptance point of view, the separator must be as open a structureas possible, safety and long-term reliability require a structure resis-tant to the perforation of conductive particles, and thus, a separatoras sealed as possible. Therefore, each separator is a compromisedepending strongly on the application for which the battery is de-signed.

Furthermore, the requirements of the separator differ between thevarious battery technologies. But in every case, some basic require-ments must be fulfilled by all battery separators. Table I gives anoverview of the requirements for the separator, their influence onbattery performance, and the solutions for fulfilling the require-ments.

The paper is structured in two parts. In the first part, the principletechniques for nonwoven production are described, and in the sec-ond part the base requirements of the different alkaline battery sys-tems and their direct influence on the separator material are dis-cussed. Because a comprehensive report covering the different

* Electrochemical Society Active Member.z E-mail: [email protected]

address. Redistribution subject to ECS terms129.2.29.132oaded on 2014-10-15 to IP

physical and chemical methods for characterizing battery separatorshas been published recently,5 no attention is drawn on this issue inthe present publication.

Nonwoven Production Technologies

The first U.S. patent dealing with nonwovens was filed in 1941claiming an artificial leather substitute.6

Before 1940 battery separators used in lead acid batteries weremade from thin wood plates, with the best performance obtained bythe use of Port Oxford Cedar wood.7 Separator developments in the1950s and 1960s included membranes, which were called dia-phragms at this time, and papers. The membranes consisted of sin-tered polyvinyl chloride, sintered polyethylene, or microporous rub-ber containing inorganic particles as pore precursors.8

Early secondary alkaline batteries used either asbestos felts or, insome instances, nonwoven materials which had been recently devel-oped and had just started to find application as battery separators.9

So even at these early times the connection between batteries andnonwoven materials had been made. Nonwoven separators havebeen commonly used as separators in alkaline battery systems sincethe 1960s.

A good overview of nonwoven materials in general and theirmain applications is given in a comprehensive book publishedrecently,10 which also shows that in terms of overall use, batteryseparators are only a very small niche application for nonwovenmaterials in general.

Nonwovens are textile products processed directly from fibers.Depending on the method of production, the fibers are used either asraw materials, or are produced in situ. Fiber materials used in bat-tery separators are predominantly synthetic polymers such as poly-olefines or polyamides, but fibers from natural origin like celluloseare also used. In contrast to woven materials, the fibers in nonwov-ens are randomly distributed; an orientated microstructure does notoccur.

The fibrous structure of nonwoven materials offers a high poros-ity, which is necessary for high electrolyte absorbance and low ionicresistance, and results in good charge/discharge acceptance of thebattery. On the other hand, the stochastic process of nonwoven pro-duction with many layers of fibers over the material’s thicknessavoids pinholes through the separator. Therefore, the risk of electri-cal shorts is minimized. This stochastic arrangement of the fibers isthe main advantage of a nonwoven material compared with wovenstructures for the battery separator applications.

Compared with membranes, the porosity of a nonwoven are gen-erally much higher. Typical membranes used as separators for sec-ondary Li batteries have porosities of about 40%, while nonwovenbattery separators have up to 75% pore volume. An increased po-rosity positively influences the electrolyte storage capability and the

) unless CC License in place (see abstract). ecsdl.org/site/terms_use of use (see

http://ecsdl.org/site/terms_use

A482 Journal of The Electrochemical Society, 154 �5� A481-A494 �2007�A482

Downl

charge/discharge capabilities. On the other hand, a common non-woven material is not a membrane. From a simple point of view, amembrane is a sheet, in which holes are introduced, while a non-woven consist of a void volume, around which fibers are laid down.The main differences between typical membranes and nonwovenseparators are presented in Table II. Note that the separators are usedin different battery systems.

The nonwoven production itself can be separated into three fun-damental steps �see Fig. 1�: formation of the web, bonding of theweb �fixing of the fibers�, and post-treatment �finishing�.

In general, web formation and web bonding are done in oneproduction step, because the nonbonded web is difficult to handle.Post-treatment is normally done in a separate step. Additional treat-ments like thickness adjustment might also be done in-line, or in aseparate step.

Formation of the web.— The technologies of web formation,which are relevant for alkaline battery separators, are dry-laid pro-cess, wet-laid process, spun-bond process, and melt-blown process.Each of these technologies is used to a greater or lesser extent in theproduction of battery separators. Depending on the battery applica-tion, one formation technology will be favored over the others be-cause of performance or fabrication advantages or possibly the cost.Dry-laid process.— Dry-laid nonwovens are produced usingcrimped staple fibers with typical lengths of 30 to 60 mm. For bat-tery separator applications fiber diameters are normally 10–20 �m,significantly finer fibers normally cannot be used in the dry-laidprocess. The fibers always contain a fiber oil �which can be a singlechemical substance or, more commonly, a mixture�, which is neces-sary for processing the fibers, for example, to guarantee their anti-static behavior. Common fiber oils used for battery separators arehydrophilic substances of different chemical nature �such as phos-phoric acid esters�.

The production process starts with the opening of the fiber bales,the different fibers are then blended in required proportions before

Requirement

Chemical resistance against the electrolyte and oxidation BLow ionic resistance �→ porous!�

High electrical resistanceGood barrier against particles

released/grown from/at the electrodesMechanical support of electrodes during cycling;

no mechanical release of active material particlesHigh absorption rates of the electrolyte �“initial wettability”�

Permanent wettability for the electrolyteHigh absorption capacity of the electrolyte �reservoir effect�Good mechanical strength

Table II. Basic differences between a separator membrane (used,for examples, in Li batteries) and a separator nonwoven (used inalkaline batteries).

Membrane Nonwoven

Structure Sheet with holes Fibers aroundempty space

Polymers Mostly PO PO, PA, andothers

Thickness �25 �m 80–300 �mPorosity �40% �50–75%Max. pore size �1 �m �10 �mPore structure Hole LabyrinthAir permeability

at 100 Pa pressure�1

L/s/m2100-500L/s/m2


being introduced to the carding machine. The carding machine is theheart of the dry-laid production process, since it determines thequality of the nonwoven and thus, the behavior of the finished sepa-rator. The principle of a carding machine is shown in Fig. 2.

In principle, the fibers are laid down onto a large cylindrical rollfrom a feeding roller. The fibers are then removed and laid downseveral times from/onto the cylindrical roll by means of worker andstripper rolls. This process ensures good mixing of the individualfibers on the microscale. Finally, the mixed fibers are removed fromthe roll by a doffer roll and are deposited in several layers onto acontinuously moving belt. Homogeneity of the nonwoven stronglydepends on the quality of the carding machine, and generally, twocylindrical rolls are combined in a series in high-quality cardingmachines.

To further improve the homogeneity, more than one carding ma-chine is used to produce dry-laid battery separators. Thereby, theproduced fleece can be orientated in the machine direction, in thecross direction �by depositing the web onto a second moving belt,perpendicular to the first one, in a zigzag pattern�, or in a combina-tion of both, depending on how the carding machines deposit thematerial onto the production belt.

One important advantage of dry-laid nonwovens as battery sepa-rators is their good initial wettability due to the presence of the fiberoil mentioned above. Thus, dry-laid nonwovens normally can beused without any post-treatment. However, it is important to estab-lish that the fiber oil is chemically stable within the battery environ-ment and will not adversely affect performance. This is especiallyimportant for secondary batteries, where the strongly oxidizing con-ditions of the charging cycle can cause decomposition of some oils,

Important for Solution

use in general �cycle life� Use of the “right” polymersBattery performancerge/discharge acceptance�

High porosity,structure “as open as possible”

attery use in general Nonconductive materialy performance �cycle life� Small pores,

structure “as closed as possible”Battery performance�cycle life, capacity�

Compressive material

Battery production Use the right polymer wetting agentsand/or posttreatment

y performance �cycle life� Use the right polymer posttreatmentry performance in general “Right” structureBattery production Strong fibers,

good bonding

Figure 1. �Color online� Schematic overview over the different steps ofnonwoven production.

Table I. Basic requirements for the separator from battery point of view and how these requirements are transferred to a nonwoven separator.

attery

�chaB

Batter

BatterBatte



A483Journal of The Electrochemical Society, 154 �5� A481-A494 �2007� A483

Downl

resulting in the formation of carbonate and other products such asmineral acid salts. Carbonate, for example, can lead to passivationof the electrodes resulting in a fading of the capacity of the cell.Furthermore, when using fiber materials, which are not intrinsicallyhydrophilic, for example polyolefines, the separator will lose hydro-philicity if decomposition of the fiber oil occurs. This “dry-out” ofthe separator can result in a reduced cycle life.

The main disadvantage of dry-laid battery separators is that theyuse relatively coarse fibers and homogeneity is inferior when com-pared to wet-laid products.

Dry-laid materials are currently used as:

1. Separators for NiCd batteries. Here, due to the electrochem-istry, the required separator thicknesses are relatively high, and thusthe need for the finest fibers and best homogeneity is not required�for details, see below�. Also, historically dry-laid separators havebeen used successfully in these applications and the need for changehas not been necessary.

2. Membrane support materials in different applications, for ex-ample, industrial NiCd batteries, primary and secondary Zn-air bat-teries, and NiZn batteries. Here, the membrane is the real separatingelement, which prevents dendrite growth. The purpose of the non-woven material is as a mechanical support and as protection for themembrane and an electrolyte reservoir and wetting aid for themembrane.

Wet-laid process.— The wet-laid process is adopted more or lessfrom paper production. The raw materials used for wet-laid nonwov-ens are noncrimped, so-called “short-cut” fibers with lengths typi-cally below 15 mm. Highly fibrillated materials like polymeric pulpscan also be used as raw materials. The raw materials are blended anddispersed in large quantities of water. The fiber concentration in thisdispersion of below 1 g/L is the main difference compared with that

Figure 2. �Color online� Sketch of a carding machine producing dry-laidnonwovens.

Figure 3. �Color online� Schematic overview of a hydro-forming machineproducing wet-laid nonwovens.


of the paper process, where relatively concentrated dispersions areused. The diluted dispersion is brought to the so-called “hydro-former,” the heart of the wet-laid process �see Fig. 3�. Here, thedispersion is sucked through a continuously moving screen belt, onwhich the fibers remain randomly distributed. Holes, which are ini-tially formed when the fiber suspension is laid down on the screenbelt, generate an increased suspension flow and thus, are immedi-ately plugged by fibers. This, together with the stochastic orientationof the fibers themselves, guarantees the best possible structure.The process additionally allows the handling of very fine fibers��10 �m diam� and produces products of high uniformity.

By using fine fibers, wet-laid materials with good homogeneityand reasonable mechanical strength can be produced in thicknessesdown to below 20 �m.11 Such materials can be used as supportmaterial in Li-polymer batteries.

The main advantages of the wet-laid process for separators arethe excellent homogeneity combined with the possibility of usingultrafine fibers. Following the tendency for increased cell capacitythe separator is required to become thinner and thinner. To excludeshorts, and thus to guarantee an acceptable cycle life, the need for afine-fibered and perfectly homogeneous separator has become moreand more important.

The main disadvantage of the process is the washing-out of thehydrophilic fiber oil during the processing, which can result in zeroor poor wettability. To obtain an acceptable initial wettability, a post-treatment is indispensable.

The wet-laid process is the standard technology for separatorsused in NiMH and Alk-Mn batteries. It is also becoming more andmore prevalent in NiCd applications.

Spun-bond process.— Spun-bond nonwovens are produced by melt-ing polymers and extruding the melt through spinning nozzles. Thismeans that the fibers are produced in situ. The so-produced endlessfibers are quenched and stretched with a hot air stream before beingcontinuously deposited onto a moving belt �see Fig. 4�. For batteryseparator applications relatively fine fibers are produced��11 �m�, but the homogeneity is not as good as with wet-laidproducts.

Because the spun fibers do not contain fiber oils, a post-treatmentto obtain initial wettability is essential.

Figure 4. �Color online� Schematic view of the spun-bond process.




Downl

In addition to battery separators, spun-bond materials are used asmembrane support materials in niche applications.Melt-blown process.— Melt-blown nonwovens are a further devel-opment of spun-bond materials in the direction of reduced fiberdiameter and improved homogeneity. Again, the polymers aremelted and extruded through nozzles. In contrast to the spun–bondprocess, a high-velocity stream of hot air is used to stretch the fibersto very small diameters ��5 �m�. Typical air stream velocities areabove 100 m/s. To obtain fine fibers, special requirements for theviscosity of the polymer melt exist. In general, a low viscosity of themelt is advantageous for the production of fine fibers. As in thespun–bond process, the fibers are deposited on a moving belt. Again,a post-treatment is indispensable to obtaining initial wettability.

The advantages of melt-blown battery separators is their highuniformity combined with low pore sizes. Both are key parametersfor separator materials. The main disadvantage is their low mechani-cal strength, which prevents their use in round cells.

Post-treated melt-blown materials are currently used in large,prismatic NiCd and NiMH batteries, where the advantages can beutilized and the requirements for a high mechanical strength is notneeded. Other applications include rechargeable button cells and asa membrane support in Zn-air button cells �see below�.

Laminates of melt-blown materials with support materials �othernonwovens or webs� can increase the mechanical strength, but alsoincrease the thickness of the whole product.

The advantages and disadvantages of the different productionmethods for battery separators are schematically listed in Table III.Of the nonwoven production methods, the wet-laid process is mostextensively used for the manufacture of battery separators.

Bonding of the web (fixing of the fibers).— After web forma-tion, the fibers are not fixed to each other, and the loose web has theappearance of cotton wool. The purpose of the bonding step is to fixthe fibers. Due to the problems associated with any handling of thenonbonded web, this bonding step is normally done in-line, just afterweb formation. Typical web bonding processes are application ofbonding agents �chemical binders�, thermo-bonding of the fibers,needling, and hydro-entanglement.

Again, all of these bonding technologies have particular advan-tages and disadvantages for producing nonwoven battery separators.These are evaluated in the following part of the paper.

Table III. (Color online) Different techniques of nonwoven productio


Application of bonding agents.— The application of bonding agentsin the form of chemical binders is a common technology to bondwebs. Thereby, the web is treated with a chemical substances, forexample acrylates �mostly in the form of dispersions, foams, or so-lutions�, which create an adhesive surface for the binding betweenthe fibers. The application of the binder can be achieved by eitherspraying �Fig. 5a� or by impregnation in a bath �Fig. 5b�.

The advantage of bonding agents is their low cost. A disadvan-tage is the need for additional chemistry, which might disturb bat-tery performance. This is more severe in secondary battery systems.Because chemistry-free alternatives exist, binders are commonly notused in separators for secondary systems. However, due to their lowprice, the use of binders is commonly used for Alk–Mn primarybattery separators.

Thermo-bonding of the fibers.— This technology does not need anyadditional chemistry to bind the fibers together. In principle, thenonbonded web is heated to a temperature where the fibers melt orbecome sticky. The fibers are then pressed together and the materialcooled down. Crossing points of the fibers form a fixed network,which possesses high mechanical strength. Heat transfer can be doneby either hot cylindrical rolls �see Fig. 6a� or in so-called air-throughdryers, in which hot air is either blown or sucked through the non-woven, which itself is held on the porous belt �see Fig. 6b�.

For this technology, it is necessary that the fiber mixture consistsof at least two components with different melting points, a “thermo-bonding component” and a “structure component.”

Battery separators typically consist of a portion of bicomponentfibers consisting of a core and a sheath component with differentmelting points. Such bicomponent fibers already contain both com-ponents mentioned above in one fiber. Polyolefin separators, for ex-ample, contain bonding fibers with a polypropylene �PP� core and apolyethylene �PE� sheath. Heating of these fibers to temperaturesabove the melting point of the PE ��120°C� leads to a sticky sur-face caused by a molten PE layer. At the contact points of the PPfibers, stable bonding points are formed. Figure 7 shows the thermo-bonding process with bicomponent fibers.

In principle, pure PE fibers can be used as bonding fiber togetherwith PP structure fibers. However, the melting of the complete PEfiber during heating generates a microhole in the nonwoven struc-ture, which increases the risk of shorts in the battery.

thods and their advantages/disadvantages for battery separators.
n me



Downl

Battery separators produced by this technology, and especiallythose using bicomponent fibers, possess excellent mechanicalstrength. The nonwoven structure is maintained during the bondingprocess, and no additional chemistry is introduced to the batterysystem.

Although the price of the bicomponent fibers is higher comparedwith single-component fibers, the technology is the most commonlyused for battery separators in secondary systems.

Needling.— In the needling process, needles equipped with smallbarbed hooks penetrate the nonbonded material at high frequencies�see Fig. 8�. The fibers are exclusively mechanically felted by thiskind of treatment. To obtain the best bonding, the needles mustpenetrate completely through the nonwoven. The needling process iscommonly used for dry-laid nonwovens, or to fix two layers ofdifferent materials together.

Because of the tendency to create pinholes within the nonwovenfabric, the needling process is seldom used for the production ofbattery separators. Furthermore, the modulus is low at low values ofmechanical tension, which means that the material stretches easilywhen a mechanical force is applied. Therefore, the handling of suchmaterials in the production process must be carefully controlled.

However, needled separators are used in some niche applicationsin large NiCd batteries and as base materials for fiber electrodes infiber nickel cadmium �FNC� batteries.

Hydro-entanglement.— In principle, hydro-entanglement is a simi-lar process to needling, where the needles are substituted by high-pressure water jets, which lead to a mechanical bonding of the non-

Figure 5. �Color online� Schematic application binders onto a nonwovenweb �a� by spraying and �b� by use of an impregnation bath. �Figure 5a isreprinted with permission from Freudenberg Vliesstoffe KG, Germany; Fig-ure 5b is reprinted with permission from EDANA, Belgium.�


woven. Typical pressures of the water jets are above 10 MPa. Theprocess of hydro-entanglement is schematically shown in Fig. 9.

Again, pinholes cannot be excluded and the mechanical proper-ties of hydro-entangled materials limit their use as battery separa-tors. In addition, the water jets can remove the fiber oil, so that apost-treatment is required to restore initial wettability.

Note that treatment of battery separators with hydro-jets is some-times used to divide so-called splittable fibers present in the non-woven material. Such fibers contain two different incompatiblepolymers. The typical cross section of such fibers is pielike, withneighboring pies consisting of the different polymers �see Fig. 10�.Applying kinetic energy in the form of the water jets splits andseparates the single pies and creates much finer fibers than the origi-

Figure 6. �Color online� Thermobonding of nonwoven materials with helpof �a� hot rolls and �b� in a cylindrical air-through dryer. �Figure 6a is re-printed with permission Vliesstoffe KG, Germany; Figure 6b is reprintedwith permission from FLEISSNER, Germany.�




Downl

nal. The more incompatible the polymers are, the lower the energyinput required to split the fibers. Again, care must be taken to pre-vent the generation of pinholes. In general, separators normally con-sist of thermobonded materials, in which the splittable fibers arealready fixed within the structure. The water-jets can then separatethe fibers without destroying the nonwoven structure. Accordingly,these materials can be considered as thermobonded rather thanhydro-entangled.

In summary, the technology of thermobonding is mostly advan-tageous for nonwovens used in secondary battery separators �see

Figure 8. �Color online� Schematic overview of the needling process. �Re-printed with permission from Freudenberg Vliesstoffe KG, Germany.�

Figure 9. �Color online� Schematic overview of the hydro-entanglementprocess using high-pressure water jets. �Reprinted with permission fromFreudenberg Vliesstoffe KG, Germany.�


Table IV�. For separators in primary batteries, which are extremelyprice-sensitive, and which have less strict requirements regardingoxidation stabilities, the use of chemical binders is standard. Allmechanical nonwoven bonding technologies bear the risk of creatingpin-holes.

Post-treatment/finishing.— After web formation and bonding,the nonwoven material has sufficient mechanical strength to behandled. In an initial post-treatment, the thickness might be cali-brated with the help of cylindrical rolls �so-called calenders�, whichmechanically press the material to the desired value.

However, the main purposes of post-treatment/finishing are toobtain a sufficient initial wettability, to guarantee a permanent wet-tability, and to create functionality, which is beneficial to batteryperformance.

Initial wettability.— Initial wettability is important during batteryfabrication, and especially during the filling of the battery�electrode–separator stacks� with electrolyte. This step is one of themost important in determining the overall speed of battery produc-tion. The higher the initial wettability of the separator, the faster theelectrolyte filling, and thus the faster the production line speed.However, high initial wetting is not always necessary because auto-matic battery production lines use vacuum filling or centrifugal fill-ing technologies, so that the requirement for high initial wettabilityis lower. Other factors must be considered as well, because the cli-

Figure 7. �Color online� Schematicbehavior of a bicomponent thermo-bonding fiber. The fiber consists of alower-melting sheath and a higher-melting core. During thermobonding,the sheath is molten and the so-produced melt acts as binder of thenonmelted core components. �Re-printed with permission from EMS-GRILTECH, Switzerland.�

Figure 10. Cross section of a splittable fiber with pielike cross section con-sisting of two different polymers.




Downl

mate of the battery production, especially humidity, has a large in-fluence. Dry conditions negatively influence the initial wettability ofmany materials.

The easiest and cheapest way to obtain a good initial wettabilityis the application of wetting agents onto the nonwoven fibers. Here,it is important to check whether the chemistry of the wetting agentdisturbs the electrochemistry of the battery. In general, the risk ismuch lower for primary systems than for secondary ones, where theseparator must resist strongly oxidizing conditions during eachcharging step. See also the comments on chemical stability of fiberoils above. An application of wetting agents can be done with thesame equipment used for binder application �see Fig. 5�.

As mentioned above, dry-laid materials possessing fiber oil onthe fiber surface generally do not need any post-treatment to obtainan initial wettability.

Permanent (durable) wettability.— While an initial wettability isrelatively easy to obtain, permanent �or durable� wettability is not.On the other hand, permanent wettability is one of the key param-eters for separators used in alkaline secondary battery systems. Ifwettability is lost during battery life, gas �oxygen� bubbles couldbecome trapped in the nonwoven pores, and, consequently, the sepa-rator “dries out”. If the battery separator looses wettability and be-comes hydrophobic there is no longer any possibility of the electro-lyte being able to penetrate the pores of the nonwovens and displacethe gas bubbles. The result is a breakdown of cycle life. Therefore,the demands for the treatment include both a long-term stabilityagainst the alkaline electrolyte and a long-term stability against oxi-dizing conditions.

Permanent wettability is less important in primary systems be-cause no gas is released to be trapped in the pores, and in Li batter-ies, in which the electrolyte is completely inert and again no gas isproduced.

For secondary alkaline batteries, a post-treatment is necessary,whenever the polymer itself is not intrinsically hydrophilic. Polya-mide separators consisting of short-chained polyamides like polya-mide 6.6 �Nylon�, or polyamide 6 �Perlon� do not need a post-treatment at all. However, the chemically more stable polyolefine-based nonwoven separators, which are standard types in NiMHbatteries, are hydrophobic by nature. Here, a post-treatment is es-sential.

Table IV. Different techniques of nonwoven bonding methods and th


It has been shown in the past that only a few post-treatmentmethods lead to a durably hydrophilic surface, capable of withstand-ing several hundreds of cycles. Treatments with low durability leadto nonpermanent effects and show a so-called aging behavior.Plasma treatment and gas-phase fluorination are proven technologiesto modify polyolefin battery separators.

Functionality.— The technologies of acrylic acid grafting and sul-fonation not only lead to a good permanent wettability, they bothhave an additional beneficial effect in batteries.12,13 Separators pos-sessing these treatments lead to a reduced self-discharge in bothNiMH and NiCd batteries.14-21 This improved performance is as aresult of the trapping of ammonia impurities on the functionalizedseparator surface. Ammonia is part of a chemical shuttle reaction,which discharges the battery. If ammonia is bonded or fixed on theseparator surface, the shuttle is interrupted and self-discharge is sig-nificantly reduced. This phenomenon is described in detail below.

As an alternative to post-treatment it is possible to produce bat-tery separators with ammonia trapping capabilities by using certainfunctional polymers as the base polymer for fiber spinning.18,22 Afterprocessing, this functionality is still present in the final product.Alternatively, it is possible to coat the nonwoven fibers with par-ticles of functional polymer to produce the same effect.18,22 Thefiber surfaces of this type of separators can be treated by the com-mon methods for permanent wettability like plasma-treatment orfluorination, and thus they do not require one of the expensive post-treatments.

All kinds of post-treatments are listed in Table V. Note that com-binations of technologies exist, e.g., using fluorination for perma-nent wettability and the application of wetting agents for improvingthe initial wettability.

In summary, the key aspects of nonwoven production technolo-gies are:

Nonwoven separators can be used in a broad variety of batterysystems.

The preferred technique for most kinds of nonwoven separatorsis a wet-laid process combined with thermobonding.

Dry-laid products are still common for NiCd separators and asmembrane support materials.

vantages/disadvantages for battery separators.
eir ad



Downl

Melt-blown materials can be used in certain large prismatic bat-teries, but fail in cylindrical cells due to their low mechanicalstrength.

The most common polymers used are polyamide for NiCd sepa-rators, polyolefines for NiMH separators, and polyvinyl alcohol/cellulose for Alk-Mn separators.

Thermobonding is the common technology for fixing the fiberstogether, where mostly bicomponent fibers are used for binding.

Wettability is of key importance for all separators. Initial wetta-bility can be achieved by wetting agents. To obtain a permanentlywettable surface, wettable polymers or a post-treatment are neces-sary. Fluorination or plasma treatment are suitable technologies toobtain a durable hydrophilic surface on polyolefines.

A bonus effect, the reduction of self-discharge, can only beachieved by chemical treatment, where sulfonic acid or carboxylicacid groups are introduced onto the polyolefine surface. Reliabletechnologies for this are grafting of acrylic acid, and sulfonationwith fuming sulfuric acid.

Alkaline Battery Systems and Their Requirements for theSeparator

Here, the main requirements of the battery are considered to-gether with how the separator can best fulfil these requirements.Often, a simple analysis of the electrochemical cell reaction givessome indication of the properties required from the separator. Forexample, thermal aspects �e.g., heat release during discharging� andsolubility of electrode materials �e.g., Zn� have a fundamental influ-ence for the requirements of the separator. Also, possible volumechanges of electrode materials during cycling must be taken intoconsideration. The main requirements of a battery separator arelisted in Table I.

In the following section the main requirements of common alka-line battery systems are described and how they are influenced bythe separator. For a good overview of the different battery technolo-gies, including the different electrochemistries, the reader shouldrefer to the standard battery literature.23,24

Please note, for easier comparison, all SEM separator photo-graphs in the following section use the same magnification �200 x�,except the figures showing a Zn-air membrane.

Alkaline-manganese battery system.— The main technical re-quirements for the separator in the primary Alk-Mn system are lowpore diameters, to avoid penetration by zinc dendrites, low weightand thickness, to maximize capacity cell, a very high initial wetta-bility, to guarantee high line speeds of battery production, and agood electrolyte absorption. Additionally, this segment of the alka-

Table V. Different techniques of hydrophilic post-treatment methods

Technique Principle Pros for separ

Wetting agents Wetting agent sits looseon fibers

Cheap,good initial wett

Coronatreatment

Application ofenergy + oxygen

None

Fluorination Gas-phase process withF2�+O2/SO2�

Very good permwettability

Plasmatreatment

Application ofenergy + oxygen

Very good initipermanent wett

Grafting ofacrylic acid

UV-induced grafting of AA�and other unsaturated compounds�

Very good initial wAdditional ef

�reduced self-disSulfonation a� Wet process with

H2SO4b� gas-phase process

�SO3�

Additional ef�reduced self-dis


line battery market is extremely price-driven, especially because theprimary batteries have to compete more and more with secondaryNiMH batteries.

The formal discharge reaction of the Alk-Mn system can be writ-ten as

Zn + 2MnO2 + 2H2O → Zn�OH�2 + 2MnOOH �1�This means that during discharge, electrolyte is continuously con-sumed. Therefore, one requirement of the separator is a high elec-trolyte storage capacity to provide sufficient electrolyte throughoutand especially at the end of the discharging process.

Technically, Alk–Mn separators consist of wettable polymers,mostly a combination of polyvinyl alcohol �PVA� microfibers, to-gether with highly fibrillated, alkaline-resistant cellulose pulp orother cellulosic components. Cellulose provides an excellent wetta-bility and, in pulp form, successfully reduces the pore size to lowvalues. Bonding is most commonly achieved using soluble PVAfiber binders. Bilayers of two light and thin nonwoven materials areadditionally used to increase electrolyte storage capability. An addi-tional membrane, for example cellophane, might be used in somecases.

Scanning electron microscopy �SEM� pictures of a typical single-layer Alk-Mn separator are shown in Fig. 11a and b.

Secondary Alk–Mn systems are on the market, but their presenceis very limited. The very low charging currents together with thelow cycle life are the fundamental disadvantages over NiCd andNiMH systems. The steady increase in capacity of common NiMHbatteries also reduces the advantage of secondary Alk–Mn batteries.

The separator in the secondary system has additional require-ments compared with the primary system. The separator must bemore resistant to oxidation, permanently wettable, and even moreprotective against dendrite penetration. Generally, the separator usedis a polyamide nonwoven in combination with a wettable membranelike cellophane. The nonwoven acts as a mechanical support/protection and also ensures that sufficient electrolyte is alwayspresent within the vicinity of the membrane to make sure that thepores are always filled with electrolyte.

Nickel–cadmium system.— In principle, the applications forNiCd batteries can be roughly classified into two groups: consumerand industrial applications. The main focus on most consumer ap-plications is low cost while high rate capability is necessary forpower tool batteries. For industrial applications, the main focus isreliability. Industrial NiCd batteries are used for many applicationsincluding emergency lighting, uninterrupted power systems, andmedical and defence systems. Each of these applications requirestailor-made batteries and often also require a tailor-made separator.

their advantages/disadvantages for battery separators.

Cons for separator Resumé

Chemistry might disturb cell performance,no permanent

wettability

Not suitable forhigh performance batteries

In air,no permanent wettability

Not suitable

Initial wettabilitycould be improved,no additional effect

Reliable technologyfor standard applications

No additional effect Reliable technology forstandard applications

lity,

�

Expensive Reliable technology forhigh-end applications

�Expensive

low initial wettabilityReliable technology forhigh-end applications

and

ator

ability

anent

al andabilityettabifectchargefectcharge




Downl

Some of the more important separator properties needed forNiCd applications are now considered. The complete chemical reac-tion of charging/discharging gives some indication of the attributesrequired of the separator

2Ni�OH�2 + Cd�OH�2 � 2NiOOH + Cd + 2H2O �2�

During charging, water �electrolyte� is produced and then it is con-sumed during discharging. This means that the electrolyte plays anactive role in the overall electrochemical reaction. Thus, the storagecapability of the cell, and more particular the separator, must be highenough to supply all the electrolyte for the electrochemical reactionwhile, ensuring that at the end of the discharge process sufficientelectrolyte still remains in the separator to allow for efficient iontransport.

Furthermore, the electrolyte is electrochemically decomposedduring charging. Especially at high state of charge �SOC� levels,oxygen is generated at the positive �nickel hydroxide� electrode as aside reaction to the main charging reaction25

4OH− → 2H2O + O2 + 4e− �3�For sealed cells it is important that the separator has sufficient

porosity to allow the oxygen generated to pass through the separator,so that it can be reduced at the negative �cadmium hydroxide� elec-trode. This process is essential to prevent excessive internal pressurebuildup with the risk of electrolyte leakage and venting on over-charge or trickle charge. Another aspect of this process is the neces-

Figure 11. Typical polyvinyl alcohol/cellulose wet-laid nonwoven materialused in Alk–Mn batteries �Style FS 22845 by Freudenberg Vliesstoffe KG,Germany�.


sity for the Cd electrode to have excess capacity to the Ni electrode.Furthermore, the reduction reaction is exothermic.26

Recent trends in power tool batteries have been in the directionof higher discharge currents and increased capacities. The higher thedischarge currents the greater amount of heat is released from theelectrochemical system. In addition, the need for higher capacitiesleads to more and more compact cells and battery packs, which aremore and more difficult to cool.

Both of the above trends lead to increased demands regarding thethermal stability of all components of the cells, including the sepa-rator. Peak temperatures in high-capacity power tool batteries mayexceed some 120°C.

At these temperatures, the standard polyamide separator materi-als undergo an accelerated hydrolysis. On the other hand, the chemi-cally more stable polyolefine materials start to melt and/or showhigh values of thermal shrinkage under these thermal conditions. Asa result, new developments of separator materials used in this kindof batteries include special polyamide materials, which combine thethermal stability of polyamide with the chemical stability of poly-olefine.

For the separator requirements, this means that:

1. The thickness need not be at very low values. Typical sepa-rator thicknesses in NiCd batteries are in the range of 150–200 �m.

2. Consequently, the need for ultrafine fibers and highest homo-geneity is not required, and typical fiber diameters are in the rangeof 10–20 �m.

3. Polymers used in applications with increased discharge cur-rents have to resist temperatures above at least some 120°C, newertrends especially in power tool applications require even higher tem-perature stability. Polyolefine-based materials might fail here.

4. Hydrolysis resistance against the electrolyte also at elevatedtemperatures is essential for high-rate discharge applications.

For industrial applications, where the focus is on reliability andcycle life performance, multilayered separators are used consistingeither of several nonwoven layers, or of nonwoven-membrane com-binations. When membranes are used, gas bubbles cannot migratethrough the separator, and, thus, such a separator combination canonly be used either in vented cells or cell types, which limit themaximum SOC to below the point where oxygen evolution occurs.

In general, the need for high-temperature stability in large indus-trial batteries is limited, where it is required polyolefine materialscan be used.

Figures 12a and b show a dry-laid polyamide separator used inNiCd power tool batteries while Fig. 13a and b show a wet-laidmaterial for a similar application.

Nickel-metal-hydrid system.— As for NiCd batteries, applica-tions of the NiMH technology are numerous. The advantage ofhigher specific energy compared to NiCd and lower price comparedto Li-ion batteries offer a broad range of consumer applications.Additional future applications arise as a result of the banning ofNiCd batteries in consumer applications.

The mostly challenging new application for industrial batteries isthe use in hybrid electric vehicles �HEVs� and as an energy buffer infuture fuel cell vehicles.27,28 Currently, the NiMH technology is theonly one that is commercially used for HEVs. The long-term reli-ability of the system is proven. However, the attempts of introducingsecondary Lithium systems for HEV batteries are on-going, but cur-rently are hindered by the not yet proven long-term reliability, thesafety issues of the Li-ion technology, and the still higher price.However, if these obstacles can be overcome, it is expected that Libatteries will play an important role in future HEV batteries, andmost likely both technologies will coexist in this application.

In the overall charging/discharging reaction of the NiMH system,the electrolyte is not involved




Downl

Ni�OH�2 + M � NiOOH + MH �4�

This means that, in principle, the separator need not act as an elec-trolyte donor or absorber. Consequently, separators used in NiMHsystems can be made thinner compared to NiCd separators. Currenthigh capacity consumer batteries use separators with thicknesses ofor below 120 �m. The reduced separator thickness is one reason forthe higher values of energy density �energy per volume� of NiMHcompared with NiCd batteries.

However, separator thickness cannot be reduced to infinitely lowvalues because it still needs to act as an electronic insulator betweenthe electrodes and also the electrolyte in NiMH systems is not com-pletely inert. At high SOC electrolyte decomposition occurs to lib-erate oxygen gas �Reaction 3�.

The basic reaction scheme is the same as in NiCd cells: oxygenis generated at the positive electrode. The released bubbles mustpass across the separator to become reduced at the negative elec-trode. This means that as long as the SOC is low, and no electrolyteis decomposed, the separator does not have to act as electrolytereservoir �see Fig. 14�. At high SOC levels, electrolyte is decom-posed and a reservoir effect is necessary.

Heat generation during high-rate discharge is less than for NiCdcells, because the main discharge reaction is endothermic.29,30 Inexperimental studies, Wu et al. found only moderate warming tobelow 50°C even at high current discharges for NiMH cells.30 Forthe separator, this means that it is not exposed to such high tempera-tures as occurs during high-rate discharge of NiCd cells. Therefore,

Figure 12. Typical polyamide dry-laid nonwoven material used in NiCdbatteries �Style FS 2647-20 by Freudenberg Vliesstoffe KG, Germany�.


Figure 13. Typical polyamide wet-laid nonwoven material used in NiCdbatteries �Style FS 22167-18 by Freudenberg Vliesstoffe KG, Germany�.

Figure 14. �Color online� Schematic overview for the relationship betweenthe charging reaction and the electrolyte decomposition side reaction on thepositive �nickel� electrode vs SOC. At high SOC levels �gray area�, theelectrolyte decomposition becomes more and more relevant, and here, anelectrolyte reservoir is necessary.




Downl

polyolefine-based nonwoven separators can be used without any re-strictions, including in high-rate discharge applications like powertools.

One of the largest drawbacks of the NiMH system is the highvalue of self-discharge, especially at elevated temperatures.14-21 Oneof the main reasons for self-discharge is nitrogen-containing impu-rities, mainly released from the nickel hydroxide. In the so-callednitrite/nitrate-ammonia-shuttle reaction, nitrate impurities in thenickel electrode are released into the electrolyte and migrate throughthe separator to the metal hydride electrode. Due to the high activityof the hydrogen atoms in the negative electrode, the anions arerapidly reduced to N�-III� in the form of ammonia. Ammonia canthen pass back through the separator and reaches the nickel elec-trode where it is oxidized back to nitrate/nitrite. The process can berepeated and both electrodes are slowly discharged.

Elevated temperatures increase the self-discharge of NiMH cellsdue to the fact that both diffusion process and chemical reactionrates increase with temperature.

In principle, this mechanism is similar in NiCd cells, where anitrite-ammonia-shuttle seems to occur.14 However, due to the lowerreactivity of oxidized species at the Cd electrode compared to that atthe metal hydride electrode, the reduction rate of the oxidized spe-cies is much lower resulting in a lower self-discharge rate for NiCdcells.14

It has been shown that separators which absorb ammoniastrongly decrease the self-discharge rate. These so-called functionalseparators are either post-treated by a grafting of acrylic acid or bya chemical sulfonation. The chemical groups introduced by thesetwo processes bind the ammonia in the strongly alkalineenvironment.19-21 It has also been proven that the introduction ofacrylic acid into the polymer prior to nonwoven processing showsan ammonia trapping effect.18,22

Unlike separators used in the NiCd system, short-chain polya-mides are not preferentially used as separator materials in NiMHsystems. These decompose slowly by releasing nitrogen impurities.It has been mentioned elsewhere that separators can be classified as�i� ammonia-releasing separators �or ammonia sources�, like polya-mides, which lead to an accelerated self-discharge, �ii� ammonia-neutral separators, for example nonfunctionalized polyolefines,which contribute neither positively nor negatively to the self-discharge, and finally �iii� ammonia-trapping separators �or ammo-nia sinks� like functionalized polyolefines, which reduce the self-discharge significantly.18

Another important issue is the control of self-discharge of singlecells in a multicell battery. In such systems, it is important that allindividual cells are on the same SOC level, even after thousands ofminicycles. This is an especially important issue in batteries used forHEVs which contain hundreds of single cells. The reduction of self-discharge by functional separators helps to equalize the SOC withinthe individual cells.

The only disadvantage of functional treatments is their high cost.Nevertheless, such functionalized separators are standard materialsin high-end NiMH batteries.

For the HEV application, additional requirements regarding qual-ity and the absence of defects must be considered. It must be held inmind that a single pinhole in the �10 square meters of separator ina HEV battery will lead to a failure of the whole battery system.Therefore, 100% control of the separator integrity is essential.

Figures 15a and b show a typical wet-laid polyolefine separatorused in round cells for consumer applications. Figures 16a and bshow a fine-fibered melt-blown nonwoven used in industrial NiMHbatteries.

Nickel-zinc system.— This system has some advantages over thecommon NiCd system, especially regarding toxicity and price of theraw materials.

The overall chemical equation of the NiZn system is similar tothat of the NiCd


2Ni�OH�2 + Zn�OH�2 � 2NiOOH + Zn + 2H2O �5�Again, the electrolyte �Eq. 6� plays an active role in the electro-

chemical process. Due to the higher cell voltage compared withNiCd and NiMH cells, the side reaction of electrolyte decompositionduring charging is more dominant, and cell design of NiZn batterieshas to be either an open or a valve-regulated one. The introductionof sealed cells for some applications was forecast some years ago,but none are currently on the market.

Nevertheless, due to technical difficulties, the system is still notwide spread, although review papers published some 10 and 5 yearsago describe a continuous increase of activities.31,32 An additionaldisadvantage is that, while NiCd and NiMH cells can be used assubstitutes for Alk–Mn, the NiZn system cannot because of the cellvoltage �1.7 V�. The main technical challenge for the separator islinked with the chemistry of zinc. The solubility of zinc in the elec-trolyte by forming zincate complexes is much higher than that ofCd, and therefore, also its tendency to form dendrites.33 These twofactors, the need of a sufficient electrolyte reservoir and the need tocontrol dendrite growth, determine the nature of the separator. Typi-cal separators consist of a combination of one or more layers ofwettable membranes and polyamide nonwovens. The membrane in-hibits diffusion of the zincate ions and dendrite penetration; whilethe nonwoven acts as mechanical support/protection for the mem-brane, an electrolyte reservoir and ensures that the membrane re-mains wettable.

Zinc-air system.— Primary zinc-air batteries are commonly usedin applications, where high specific energy and energy density are

Figure 15. Typical polyolefine wet-laid nonwoven material used in NiMHround cells �Style FS 2225-12 N20 by Freudenberg Vliesstoffe KG, Ger-many�.




Downl

required. Because one of the two active materials, oxygen, is sup-plied from the environment, the majority of the cell space can beutilized by the zinc anode.

Therefore, Zn-air batteries are used in hearing aids, medical sys-tems, and in back-up and defence applications, where the highestvalues of specific energy are required. Another niche application isthe use of Zn-air batteries for electrical fences. Due to the hightendency of zinc to form dendrites, special nonwoven-laminatedpolyolefine membranes are used as separators. These membranes aretreated with certain wetting agents to guarantee the wettability of thepores. Figures 17a and b show a separator material �laminate ofmembrane and melt–blown nonwoven� used in Zn-air batteries. Anadditional nonwoven layer can be used as extra support or to act asan additional electrolyte reservoir.

Because the overall discharging reaction does not incorporateelectrolyte, the separator membrane can be relatively thin. However,the electrolyte is not inert: the first step of oxygen reduction in-volves water, which is supplied from the zinc anode. Therefore,Zn-air membranes are not as thin as those used in lithium batteries�where the electrolyte is completely inert�. Thicknesses of commonZn-air membranes are 80–100 �m

Zn + 12 O2 → ZnO �6�

Another function of the membranes or/and nonwovens is to con-trol the access of oxygen into the cell. The incoming air has to passa hydrophobic membrane, commonly consisting of Teflon. The hy-drophobic nature of the membrane prevents additional water being

Figure 16. Typical polyolefine melt–blown nonwoven material used in in-dustrial NiMH batteries �Style FS 2192-11 SG by Freudenberg VliesstoffeKG, Germany�.


introduced into the cell. The gas then passes a gas diffusion non-woven material, which distributes the air evenly over the conductivecatalyst �commonly MnO2 on activated carbon�, where the oxygenis reduced.

One problem for normal consumer applications is the extremelyhigh rate of self-discharge of activated Zn-air batteries. For storageand before use, the gas inlet of the batteries is sealed by a foil, whichguarantees a reasonable shelf-life. The foil must be removed prior touse to activate the cell. Once opened, self-discharge is high. Avoid-ing this problem, and thus drastically increasing the shelf-life of theactivated batteries, would be a great advance for the Zn-air technol-ogy.

Another important technical and environmental issue is the mer-cury �Hg� additive in the zinc powder �ca. 1%� necessary to suppressgassing. While Alk–Mn consumer batteries are now produced Hgfree by the use of pure raw materials,34 mercury cannot currently beavoided in Zn-air batteries.

Secondary zinc-air batteries are a much more problematic systemfor several reasons �see below�. Nevertheless, there are a number ofindustrial uses for this battery system, focused on large-scale batter-ies for traction and back-up applications. The batteries are eithermechanically �by replacing the oxidized Zn products by fresh Znmetal�, or electrochemically rechargeable.35 The main advantages ofthe Zn-air system are the low costs of the raw materials together

Figure 17. Laminate of membrane and melt–blown nonwoven material usedin Zn-air batteries �Style 5550 by Celgard, USA�. �a� The cross section withthe membrane on the top. Note that the binding between the two layers isdone by a point-sealing, which is not visible in the picture. �b� A top-viewonto the membrane surface. Note that the magnifications of the two pictures�500x for the cross section, and 10,000x for the top view differ from the onesin the previous pictures�. �Reprinted with permission from Celgard.�



echnic


Downl

with the nontoxicity of Zn. However, if Hg is needed to suppressgassing, these perceived advantages may be of limited value.

Detrimental from the technical point of view are the high zincsolubility in the electrolyte and its tendency to form dendrites. Ad-ditionally, the uptake of CO2 coming from the air into the electrolyteby forming carbonate is a major concern. Carbonate can then passi-vate the zinc electrode by forming insoluble ZnCO3. While this isnot an issue for primary batteries, CO2 enrichment increases withthe number of cycles. For this reason, a gas-separating membrane inthe air inlet preventing CO2 from entering the cell might be ofinterest. However, the current gas-separating membranes are muchmore permeable to CO2 compared to oxygen, so an easy filtering ofthe CO2 currently is not possible. The removal of CO2 in a by-passmight be a possibility, but would require additional energy to gen-erate a pressure difference to evacuate the CO2 from the main airstream. An “intelligent” removal of the CO2 could be a subject offuture work.

Finally, the shape change of the Zn electrode during the multiplecharging/discharging cycles must be taken into account. Unlike, forexample, the cadmium electrode, the deposition of Zn during charg-ing is less controlled, and thus the tendency to release loose particlesis increased.

In general, the separator in the large batteries must consist of amembrane together with a nonwoven support material, which is nec-essary to mechanically protect the membrane, to fix the anode ma-terial, and to guarantee a wetting of the membrane with electrolyte.For secondary applications, the need for a permanently wettablemembrane is urgent, because materials, which are only treated bywetting agents, might fail prematurely during the cycle life.

Due to the unresolved problems mentioned above, the Zn/airtechnology is currently not commercially used for large-scale sec-ondary batteries, despite efforts in the past to use it, particularly fortransport applications.

In Table VI the characteristics of the commonly used recharge-able battery separator used in NiCd, NiMH and Li-ion are com-pared. The separator thickness can be reduced with increased inert-ness of the electrolyte in the electrochemical system. This is thereason for the very low thicknesses of Li-ion battery separators.Additionally, gas transport through the separator in sealed alkalinesystems requires an open structure with pores which are not toosmall. Membranes cannot be used in such systems. On the otherhand, the tendency of particle penetration through the separator andthe formation of dendrites during over-charging are very high in theLi-ion system. For this reason, membranes with small pores must beused. However, the current membranes possess relative low porosityvalues of �40%. It might be an object of future work to developnonwoven separators possessing a thickness comparable to that ofmembranes, while still maintaining the high porosity of the non-woven material.

Outlook for Future Developments

Trends for future battery separators follow at least two direc-tions:

Table VI. Comparative overview of the three common rechargeablrequirements for the battery separator. The partial inertness of the elin Li-ion batteries are reasons for increased specific energy of these

Battery system Electrochemistry Com

NiCd Electrolyte is reaction partnerduring charging/discharging

Thickness cato “lo

NiMH Electrolyte is consumedonly during charging at

high SOC levels

Thickness cbut no

Li-Ion Electrolyte completely inert Thicknessto “lowest t


1. For consumer applications, price will be the dominant param-eter. It will have to be a major consideration in future research.

2. For high-end applications, the need for tailor-made batteryseparators will become more and more important.

The key areas for future development are therefore reduced cost,tailor-made separators, long-term stability �HEV�, reduced self-discharge �competition to Li-ion�, and safety.

Conclusion

This paper has shown that there are many different requirementsfor a separator used in the different alkaline battery technologies andthere are many different technologies for producing nonwoven bat-tery separators. Combining the right nonwoven technology with theexact requirements of the battery system is the key issue to obtain-ing the best possible separator.

Acknowledgments

The authors acknowledge the support of our colleagues MichaelAppelgrün, Hans Feistner, and Alexander Koch for their help cor-recting the manuscript, Christlorenz Bundi from EMS GRILTECH,Switzerland, for providing the pictures of thermobonding fibers,Christian Schäfer from Fleissner, Germany, for providing the pic-tures of an air-through dryer, Pierre Wiertz from EDANA, Belgium,for providing the picture of the impregnation processes, and JürgenSchneider from Celgard, Germany, for providing samples of thematerial Celgard 5550.

Freudenberg Vliesstoffe KG assisted in meeting the publication costs ofthis article.

References1. G. C. Zguris, J. Power Sources, 73, 60 �1998�.2. G. C. Zguris, J. Power Sources, 107, 187 �2002�.3. W. Böhnstedt, J. Power Sources, 158, 1102 �2006�.4. P. Arora and Z. Zhang, Chem. Rev. (Washington, D.C.), 104, 4419 �2004�.5. E. Guasp and M. M. Manzo, Test Procedures for Characterization, Evaluating, and

Managing Separator Materials Used in Secondary Alkaline Batteries, NASAMemorandum 107292, The National Aeronautical and Space Administration,Washington, D.C. �1997�.

6. L. C. Nottebohm, U.S. Pat. 2256034 �1941�.7. R. G. Robinson and R. L. Walker, in Batteries, D. H. Collins, Editor, p. 15, Per-

gamon Press, Oxford, UK �1963�.8. M. Barak, in Electrochemical Power Sources, IEE Series 1, M. Barak, Editor, p.

151, The Institution of Electrical Engineers, London, UK �1980�.9. H. N. Steiger, P. B. Pinches, H. I. Cohen, and R. C. Shair, in Proceedings of the

16th Annual Power Sources Conference, Atlantic City, NJ, May 22–24, 1962, p. 96.10. W. Albrecht, H. Fuchs, and W. Kittelmann, Nonwoven Fabrics. Raw Materials,

Manufacture, Applications, Characteristics, Wiley-VCH, Weinheim, Germany�2003�.

11. P. Kritzer, J. Power Sources, 161, 1335 �2006�.12. �a� J. Cook and I. Smith, Coating, 1, 19 �2003�; �b� J. Cook and I. Smith, Coating,

2, 61 �2003�.13. K. Iwasaki, M. Ikoma, M. Ikeyama, H. Kawano, and I. Matsumoto, Euro. Pat. EP

316916 B1 �1988�.14. G. A. Levina, Z. P. Arkhangel’skaya, T. M. Bessonova, E. N. Varypaeva, and N. O.

Volkova, Khim. Istochniki Toka, 23, 85 �1987�.15. M. Ikoma, Y. Hoshina, and I. Matsumoto, J. Electrochem. Soc., 143, 1904 �1996�.16. P. Leblanc, P. Blanchard, and S. Senyarich, J. Electrochem. Soc., 145, 844 �1998�.17. D. Singh, T. Wu, M. Wendling, P. Bendale, J. Ware, D. Ritter, and L. Zhang, Mater.

tery systems concerning basic electrochemistry and following theiryte in the NiMH system, and the complete inertness of the electrolyte

s compared with NiCd.

tsThickness and pore sizes

of separatorRequirements for anonwoven separator

be reducedlues

Th:200 �mP:30 �m

Homogeneity: good,fiber size: low

reduced,ero”

Th: � 150 �mP:15 �m

Homogeneity: excellent,fiber size: very low

reducedal value”

Th: � 25 �mP: � 1 �m

Nonwoven not possible→ membranes

e batectrolsystem

men

nnotw” vaan be

t to “z

can be




Downl

Res. Soc. Symp. Proc., 496, 25 �1998�.18. P. Kritzer, J. Power Sources, 137, 317 �2004�.19. J. A. Cook and I. M. Lancaster, in Selected Battery Topics, G. Halpert, M. L.

Gopikanth, K. M. Abraham, W. R. Cieslak, and W. A. Adams, Editors, PV 98-15,p. 55, The Electrochemical Society Proceedings Series, Pennington, NJ �1998�.

20. K. Shinyama, Y. Harada, R. Maeda, H. Nakamura, S. Matsuta, T. Nohma, and I.Yonezu, Res. Chem. Intermed., 32, 447 �2006�.

21. K. Shinyama, Y. Harada, R. Maeda, H. Nakamura, S. Matsuta, T. Nohma, and I.Yonezu, Res. Chem. Intermed., 32, 453 �2006�.

22. W. Wenneis, H. Hoffmann, B. Severich, and H. Keul, Euro. Pat. EP 1291938 A3�2005�.

23. Handbook of Batteries, 3rd ed., D. Linden and T. B. Reddy, Editors, McGraw-Hill,New York �2002�.

24. Handbook of Battery Materials, J. O. Besenhard, Editor, Wiley-VCH, Weinheim,Germany �1998�.


25. C. Cha, J. Yu, and J. Zhang, J. Power Sources, 129, 347 �2004�.26. S. Gross and J. Malcolm, in Power Sources 4, D. H. Collins, Editor, p. 257, Oriel

Press Ltd., Newcastle upon Tyne, UK �1973�.27. O. Bitsch and G. Gutmann, J. Power Sources, 127, 8 �2004�.28. U. Köhler, C. Antonius, and P. Bäuerlein, J. Power Sources, 127, 45 �2004�.29. T. Araki, M. Nakayama, K. Fukuda, and K. Onda, J. Electrochem. Soc., 152,

A1128 �2005�.30. M. S. Wu, Y. Y. Wang, and C. C. Wan, J. Power Sources, 74, 202 �1998�.31. J. Jindra, J. Power Sources, 66, 15 �1997�.32. A. K. Shukla, S. Venugopalan, and B. Hariprakash, J. Power Sources, 100, 125

�2001�.33. H. Chang and C. Lim, J. Power Sources, 66, 115 �1997�.34. M. F. Almeida, S. M. Xará, J. Delgado, and C. A. Costa, Waste Manage., 26, 466

�2006�.35. O. Haas, S. Müller, and K. Wiesner, Chem.-Ing.-Tech., 68, 524 �1996�.



Nonwovens as Separators for Alkaline Batteries

Documents

Transcript of Nonwovens as Separators for Alkaline Batteries