Factors Affecting the Selectionand Performance

of Silicone Release Coatings

By Darrell JonesScientist

Dow Corning Corporation

2

About the Author

Darrell Jones

Darrell Jones is the epitome of a Dow Corning “Scientist” – a visionary, a creator,an innovator of release coating technology.

In the early ‘70s at Dow Corning Europe, Darrell created the industry’s firstsolvent-based platinum-catalyzed release coatings. In the U.S., he has been animportant player in the development of solventless silicone release coatingtechnology. He has also advanced more efficient and precise solventless coatingmethods.

For many release coating customers around the world, Darrell Jones IS DowCorning. Knowledgeable. Accessible. And eager to share his problem-solvingexpertise.

For over 35 years, Darrell has been a key member of the Dow Corning scientificcommunity. What follows is one small “chapter” of Darrell Jones’ “book” ofsilicone release coating technology. For those who know Darrell, this paper willread like a conversation with a trusted friend.

© 1997 Dow Corning Corporation. All rights reserved.

3

INTRODUCTIONRelease coatings are generally used toprevent things from sticking together! Thissimplistic statement embraces a wealth oftechnology and a global industry involvingboth silicone and nonsilicone materials. Attheir simplest level, fluids or powders arecoated or sprayed onto a surface to providea nonstick surface for another materialsubsequently brought into contact with theanti-adherend. This type of release agent isnon-permanent and usually results insignificant transfer of release agent to thereleased surface. This transfer is calledmigration or offset. Migratory coatingsinclude soaps, oils, fluorocarbons, siliconeoils and even, under the rightcircumstances, water. They are effectivebecause they form a weak boundarybetween the two surfaces and result inrelease agent being lost from the originaltreated surface, such as a mold, and beingtransferred to the released surface assurface contamination. We shall notconcern ourselves with such release agentshere.

A somewhat more sophisticated set ofmaterials is semi-permanent releasecoatings. These result in some transfer tothe released surface, but it is a muchsmaller proportion. The release agent,itself, is not usually a simple fluid but acompounded material involving solidmaterials dispersed in fluids or oils. Semi-permanent coatings include waxdispersions, mixtures of silicone resins insilicone fluids, micas in various fluids andorganic resins such as acrylics.

Temporary and semi-permanent coatingsare only mentioned to highlight thespecialized nature of our subject, which isessentially nonmigratory coatings that arepermanent (or at least capable of multipleuses) and have excellent anti-stickproperties. These materials are dominatedby silicones, but some organic materials,such as urethanes, fluorocarbons andacrylics, can be used when eithertoughness of the release surface or absenceof silicone is of greater importance thanabsolute ease of

release. In this treatise, we are onlyconcerned with silicone release coatings.Silicones in many forms offer excellentrelease properties, but to present anonstick surface and, at the same time,offer little or no transfer to the releasedsurface, we need to form a cured filmicsurface free from migratory species. Thispresents many challenges, and we shallexamine various approaches used to meetthem.

Polydimethylsiloxanes have severalunique properties. Low surface energyfrom the dimethyl groups and the greatflexibility of the siloxane backbone arethe two that make the formulation ofrelease coatings possible. Silicones, orpolydimethylsiloxanes (PDMS for short),offer a dimethyl surface to the air. Thisdimethyl surface has a very low surfaceenergy. A convenient measure of this issurface tension; for PDMS, surfacetension is 21 to 22 dynes/cm. Lowsurface energy alone, however, does notguarantee a good release surface. Thesilicone backbone is an extremely flexiblepolymer chain with virtually unhinderedrotation about all the Si-O-Si bonds. Thisflexible backbone, together with lowsurface energy, are the essential elementsof silicone’s unique release capability.However, even these two properties areinadequate to give a nonmigratory releasesurface. To achieve that goal, we mustfind some way to tie all the polymerchains together in a coherent film. Thisinvolves the use of crosslinking, orcuring, chemistry, which brings us to thecrux of this discussion of curable siliconerelease coatings.

RELEASE COATINGMARKETSAs the name implies, release coatings areemployed to render surfaces anti-adhesive. By far, the biggest single use isto make release liners that carry andprotect pressure-sensitive adhesives(PSAs for short) until the moment of use.For this application, the most importantsubstrate is paper. PSA-coated materialsinclude labels, tapes, over-laminatingfilms, sign lettering,

medical devices and a host of smallerapplications. The release liner’s role is toprotect the adhesive until it is at the pointof application where it must peel awaycleanly with little use of force. This basicfunction is true of all PSA applicationswhether the liner is discarded, as in alabel or decal, or an integral part of thepressure-sensitive product such as a self-wound tape.

Other markets for release liners that arenot pressure-sensitive adhesive-basedinclude carriers for oily and sticky massessuch as sealants, gaskets and mastics;interleaving layers for rubber processing;and support and release base for thecasting of plastic films such aspolyethylene, polyurethane, acetate andvinyl films. Liners in this last group areoften embossed to impart texture to thecast film. Further examples of releaseliners include food processing aids suchas baking or pan liners; food packagingmaterials such as gum wrappers; asphaltpackaging, in the form of coated druminteriors; release-coated inner-layers ofpaper sacks; and caul or separating sheetsfor the production of high-pressure plasticlaminates.

One final example of significance is therelease liner for tiles and shingles. Thisapplication is, in essence, a pressure-sensitive adhesive application but differsfrom the earlier examples in that oneneeds a fairly firm bond between therelease sheet and the adhesive to preventpremature removal of the liner. This isparticularly important as tiles andshingles are stacked for storage, and anyexposure of the adhesive leads to the tilessticking together.

The end uses of these PSA constructionsare as varied as the constructionsthemselves. PSA products are used in themedical, construction, automobile,graphic arts, electrical and electronicindustries. Electronic data processingconsumes vast quantities of pressure-sensitive labels, most of which turn up inthe average household as address labelson bills and advertising material.Supermarkets also use significant

4

quantities of pressure-sensitive labels asprice-weigh labels, special-offer labels and,increasingly, in the form of a complexlabel that also includes a discount coupon.Most of today’s special mail and packagedelivery operations use pressure-sensitivebar code labels as check systems tomonitor the progress and delivery ofpackages. Labeling systems are beingdeveloped to provide hospitals withpermanent records of patient treatment andprogress. Most of today’s tamper-resistantpackaging is based on some form ofpressure-sensitive labeling system. The listis endless and grows daily as greater andgreater innovation is displayed by thepressure-sensitive industry. Ten to 15years ago, it was generally concluded thatabout five percent of the labeling industrywas based on pressure-sensitive materials.In 1995, that figure was estimated at 45 to50 percent.

A similar state of affairs exists in the tapefield. The most familiar tapes are officeand household tapes based on cellulosic orclear plastic films. These do not normallyinvolve the use of silicone releasematerials. However, tapes have a verywide range of capabilities and uses. In thepackaging field, greater strength in bothsubstrate and adhesive is required, andthese more aggressive adhesives donecessitate the use of a release coating.Depending on the adhesive type and therequirements of the application of the tape,this release coating is often silicone based.Electrical insulating tapes, wrapping tapes,thermal insulating tapes and decorativetapes are all areas where silicone releasecoatings find use.

In the non-pressure-sensitive field, inaddition to those applications alreadygiven, some newer specialized markets aredeveloping. One of these is for carbon-fiber/epoxy laminating materials.Although silicones used in this field havehad some problems related to siliconemigration, silicone release coatings can bemade that perform satisfactorily in thisdemanding application and provide a

stability of release not found with othermaterials. Another growing use is ananti-soil treatment for surfaces wherewater, oil and dirt repellency is required,such as high-quality decorative labels.

In considering these diverse markets andapplications, both pressure-sensitive andnon-pressure-sensitive, we must also keepin mind that the release coating is usuallya very minor part of the construction –approximately 1/2 to 1/4 of a percent byweight. The other components have avery wide range of material choices, also.The release liner, upon which the siliconeis coated and cure, can be paper, plastic,plastic-coated paper or metallic. Thepaper itself can also vary widely, rangingform dense types such as glassine toporous materials such as machine-finished kraft. Clay-coated and othercoated papers are also widely used.Plastic substrates include polyethylene ofvarious types, polypropylene, poly-halogenated olefins, polyester and acetatefilms. Metallic films and paper are alsoencountered as release liner bases.

Different liner materials have a profoundeffect on the choice of silicone, both inthe nature of the curing chemistry and onthe physical form of the coating. Ofequal impact is the nature of the adhesivemass to be released.

PRESSURE-SENSITIVEADHESIVESThis subject is covered in great detail inmany publications, and it is not my intentto delve into it here. However, when onetalks of release coatings, one cannotignore the fact that it is the totalconstruction that is under consideration.In the case of pressure-sensitive adhesivematerials, the silicone is about 1/15 to1/20 the mass and thickness of theadhesive, and it is the adhesive thatdominates the performance.

Factors affecting release will be discussedlater, but for now, suffice it to say that

rheological properties are most critical.These rheological properties arecontrolled by the chemical nature,molecular weight and formulatedcomposition of the adhesive. Cost andperformance requirements influencechemical nature, so acrylic, SBR, naturalrubber and other synthetic rubber andepoxy-based materials are commonalternative adhesive chemical types.Solvent-based, water-based and 100percent solids or hot melt are thealternative delivery systems. This has amajor influence on the molecular weightof the adhesive materials used. Finally,adhesives are asked to perform undervaried conditions such as the normal tohigh temperatures encountered in shopwindow displays and the lowtemperatures to which freezer labels aresubjected. Adhesives may be required tobe permanent or easily removable withoutdamage to the labeled article. In the tapefield, it is often desirable to remove thetape without leaving adhesive behind.

All these adhesive requirements will alterthe nature of the total pressure-sensitiveconstruction and, in turn, demand a widerange of release behavior from thesilicone. To achieve that range ofperformance is the challenge beforesilicone release coating technologists.Consider the means at their disposal:

CURING CHEMISTRYAs stated earlier, in addition to releasingmaterials in a controlled fashion, a releasecoating is really only useful in thepressure-sensitive industry if theperformance o the released adhesive isunimpaired by silicone transfer. Tominimize silicone transfer or migration,the silicone must be tied together in acured, coherent film covering all thesupporting substrate or liner. Maximumrelease performance is obtained frommaximizing the dimethyl content of thecured film. There is the beginning of aconflict that persists in all release-coating

5

formulations, even today. Speed of cure isbest served by a high concentration ofcuring groups and catalysts, whereasrelease performance is enhanced by amaximum concentration of dimethylsiloxane groups.

The following curing chemistries all hadadvantages and disadvantages.

Example A: The advantages of thischemistry are that it is very fast when asuitable catalyst is used. It is alsoinexpensive and easily made intopolymers. This chemistry is notsusceptible to poisoning, so it is unaffectedby potential substrate inhibition. Thedisadvantages include great difficulty inputting the SiOH anywhere except on theterminal silicon atoms of polymers. Thissystem needs a catalyst, normally anorganotin salt. Because it is not easilypoisoned, it is very difficult to control bathor working life once catalyzed. Hydrogenevolution can be a problem under certaincircumstances.

Example B: This is the condensationreaction and clearly could also occur inExample A. It is a fast reaction, but whenthe silanol groups are only on the terminalgroups of polymers, it is not as fast asExample A. Because it is difficult to builda polysilanol-functional molecule, it isdifficult to crosslink a curing film, and thisreaction, on its own, is not a good basis tomake cured films. The use of organotinsalts accelerates the curing, but the basicreaction is reversible. This is not desirablebecause of the potential to createmigratable species.

Example C: This chemistry has severaladvantages. Polyfunctional –ORmolecules are easily made as tetra-, tri-

and di-alkoxy silanes. A wide variety ofalkoxy and acyloxy groups is available.Within this range, cure speed varies fromvery fast to very slow. Unfortunately, themore common alkoxy groups, such asmethoxy, ethoxy, etc., are slow curing.More complicated groups, such asaminoethoxy or alkoxy groups containingeither linkages, are very fast but are alsovery hydrolytically unstable, makingthem somewhat unstable, making themsomewhat unsuitable as groups forpolymeric molecules. However, whenused with silanol-terminated polymers,these alkoxy groups can form excellentcrosslinking materials either as oligomersor silanes. The chemistry benefits fromthe use of a catalyst, but it is difficult tocontrol the bath life.

When the –OR group is an acyloxygroup, catalysts are not necessary; but theliberated acid, such as acetic acid fromacetoxy group functionality, is a definiteproblem in terms of its effect on thesubstrate and machinery and its volatilityand effect on adhesive mass.

Example D: This is a variation onExample C and uses moisture as areactant. This chemistry is the basis ofsome sealant technologies and can beused in release coatings, especially whenused with a titanate catalyst. However,bath life and cure speed are not good, andthis option has been widely used.

Example E: This chemistry is verydifferent from the foregoing examples.This system needs a catalyst, normally anorganoplatinum complex. With thiscatalyst, the system is easily poisoned orinhibited. This is a clear disadvantagewhen the substrate inhibits cure, but wheninhibition is used in a controlled manner,a long working bath life can be obtained.This is especially advantageous whenconsidering high solids or even 100percent solids materials.

Example F: Normal polydimethyl-siloxanes can be cured into rubber-likematerials through the use of peroxidecuring agents. In theory, this is a verydesirable system with the maximumamount of dimethyl siloxane. In

practice, several difficulties proveoverwhelming. First, this kind ofchemistry is inhibited by oxygen.Release coatings are virtually all surface,the film being only about 1 micron thick,so surface effects such as oxygeninhibition have a dominating effect.Second, to ensure no free polymercontent, which would lead to migration orsilicone transfer, a high loading ofperoxide is necessary. This leaves acontaminant with no release properties inthe cured film, offsetting the gain indimethyl content. Finally, the curingtemperature for such peroxide systems isquite high and damages the substrates,especially heat sensitive materials.Although this chemistry is widely used inthe silicone elastomer industry, it is notused for release coatings.

This list is not exhaustive but does coverthe most obvious alternatives facing therelease coating technologist. In the earlydays of the technology, Example A wasthe chosen technology. Polymers weremade with silanol end groups, andcrosslinkers were made from poly-SiH-functional oligomers. Catalysts such asdibutyl tin di-2-ethylhexanoate were usedat levels of 2 to 15 percent by weight.

In later years, the chemistries inExamples C and D were employed toaccelerate the cure rate of Example A.Both siloxane oligomers and silanes weremade with very reactive alkoxy ligands.These reacted with both silanol from thepolymer and water from the atmosphereand from the paper to producepolysilanols, which are very reactive andaccelerate the cure substantially.

6

Figure 1. System schematic. Figure 2. Addition-curing mechanism.

The schematic for this system is shown inFigure 1.

Control of release performance, which willbe discussed in more detail later, involvescontrolling the elasticity, or crosslinkdensity, of the cured silicone film. Withthis chemistry, this is easily achieved bycontrolling the chain length of the polymer.Polymers ranging from 20 to 5000 siloxaneunits, were employed. Because of thedifficulty of controlling this set ofreactions, working bath life could only beachieved by working in dilute dispersions,either as solutions in solvents or as oil-in-water emulsions. One hundred percentsolids operations were virtually impossiblewith this chemistry, which comprises theso-called condensation or tin-catalyzedsystem still widely used today.

The advantages are fast, low-temperaturecuring, relatively wide range of releaseavailable through polymer molecularweight control, and freedom from substrateinhibition. The disadvantages aresomewhat lengthier. There is no 100percent solids alternative; indeed, even insolvent it is difficult to operate much above8 percent solids. The reversible nature ofExample B chemistry leads to aphenomenon called blocking in which linercoated with silicone on both sides tends tobond to itself on storage. The tin catalystsare used at

quite high loadings and can alter theadhesive behavior on aging. Migrationtends to be more prevalent with thischemistry.

These coatings were the basis of thesilicone release liner business form thelate 1950s until the mid 1970s. At thattime, the environmental pressure onsolvent emissions made the developmentof 100 percent solids coatings aparamount need.

To meet this need, the addition-curingchemistry shown in Example E becamethe preference. The addition of a siliconhydride group across a vinyl group iscatalyzed by a few parts per million ofplatinum metal in the form of an organoor organo-silicon platinum complex. Thiscontrasts sharply with the 2 to 15 percentof organotin salt used in the condensationsystem. Being an addition reaction, thisreaction has no leaving groups, so nothingis evolved during curing. The curingmechanism is shown schematically inFigure 2.

The presence of another group that caneither compete with or exclude the vinylfrom the platinum is called inhibition inthe first case, and poisoning in thesecond. The difference is purely one ofdegree and desirability. In the presenceof the catalyst, the addition goes veryrapidly with the evolution of heat. Indilute solution, the bath life can be

controlled much as in the condensationsystem, but for high solids or 100 percentsolids, selective inhibition must be used.

Commonly used inhibitors effectivelyexclude the vinyl groups from theplatinum at temperatures around roomtemperature. At elevated temperaturesabove 80 °C for example, these inhibitorseither evaporate or react with the SiH andbecome part of the cured matrix. In eithercase, they can no longer exclude thepolymer vinyl groups, and cure takesplace rapidly. Such inhibitors includevarious acetylenic alcohols, materialscontaining conjugated double and triplecarbon-carbon bonds, various unsaturateddicarboxylic acid esters and someketoximes. (See Figure 3). Materialssuch as amines, sulphides, phosphidesand the organotin salts used as catalysts inthe condensation system from somewhatstronger bonds with the platinum and aregenerally regarded as poisons. Theirpresence makes it difficult, if notimpossible, to cure an addition-curingcoating.

A further highly advantageous feature ofthis chemistry is the ease of polymermanufacture. While silanol groups areonly easily placed terminally, vinylgroups may be placed not only terminallybut anywhere along the polymer chain,virtually at will. Furthermore, thesepolymers will stand

7

Figure 3. Inhibitor examples.

heat treatment better than silanol-functional polymers and so are more easilydevolatilized. This further aids the causeof migration-free coatings.

Thus, in addition to the ability to controlthe elasticity of the cured film by polymerchain length is the scope to do so bymanipulating both the number anddistribution of vinyl groups along thechain.

This chemistry has no reversible reactionsin the mechanism and so is free from theblocking phenomenon discussed earlier.There are two major disadvantages. One isthe sensitivity to poisoning either by thesubstrate being coated or by theintroduction of foreign material prior tocoating. The second Is the inherent cost.Both the catalyst and the vinyl moiety aremore expensive than the correspondingcondensation-based materials.

The rapid rise in 100 percent solidsoperations over the past 15 years has meantthat economics of scale have offset someof the inherent cost penalties, and a generalincrease in awareness of the nature ofsubstrate inhibition has rendered much ofthe sensitivity a minor concern. In theyears since the mid-1980s, the addition-cured materials have taken an increasinglybigger share of the market from the oldercondensation materials. In 1990, the sharewas greater than 65 percent on a globalbasis. The use of condensation

systems continues to decline.

Although these two systems havedominated the chemistry of siliconerelease coatings, all the examples givenand some others can serve to meet thebasic need, which is to form a cured filmof silicone in which little or no loose,migratory or extractable material remainsafter curing. The curing mechanism isnecessary to do just that and has no othervalue. Should a way be found to make athermoplastic release coating, forexample, this curing mechanism wouldno longer be necessary. Later, under“Future Developments,” radiation curingwill be discussed. This, too, Is but ameans to an end. In the release coatingindustry, the failure of the silicone tocompletely cure has been the cause ofsignificant problems. At present,however, the cure mechanism of a releasecoating is perceived to be at least asimportant as any other property of thesystem.

COATING SYSTEMSIn the early days of silicone releasecoatings, specialty applications were thenorm, and raw material costs were not anoverriding concern. As everyday usesbecame more and more common, costcontrol became more pressing. Althoughthe silicone is only about 1/2 to 3/4percent of the weight of the construction,it can be 2 to 6 percent of the cost of aplain pressure-sensitive label and as muchas 30 percent of the cost of some siliconeliners. Therefore, the desirability of usingas little silicone as possible is certainlyunderstandable. The minimum amountnecessary will vary from substrate tosubstrate. Plastic films, with their perfecthold-out and smooth surface, need onlyabout 0.1 to 0.2 grams per square meter(gsm) to effect perfect film coverage andexhibit good release properties. Aparchment, on the other hand, may almostunavoidably require 5 gsm as the siliconefills up the surface imperfections.European glassines typically need about0.7 gsm, whereas North American super-calendered kraft needs a little more, about

1 to 1.2 gsm, for complete coverage.

A coating of 1 gsm is almost exactly 1micron (one millionth of a meter) thick.The task of applying such a thin coatinguniformly and consistently has been quitea challenge to coating equipmentmanufacturers.

Coating methods can be classified inmany ways. One rough division is intomethods that apply an excess and thenmeter off to the required amount, andmethods that deliver a preciselycontrolled amount directly onto thesubstrate. Examples of the first methodare Mayer rods, air knives and blade-over-roll coaters. The best knownexample of the second is gravure.

In the early days of siliconizing, whenonly silicone solvent dispersions or water-based emulsions were available, all ofthee methods were employed. In general,even then, slot-die extrusion, foamcoating and spray coating were notsuccessful for a variety of reasons, butpredominantly because of the lowrequired coat weight.

Of the accepted methods, direct gravure isthe most accurate and the mostsuccessful, but Mayer rods and air kniveshave been used quite widely, especiallywhere slightly higher coat weights areacceptable. All three methods offer twoopportunities to control coat weight. Thefirst is the equipment itself – gravure cellsize, Mayer rod number, and air knifepressure and impact angle, all of whichhave a major effect on deposited coatingweight. A second opportunity is in thesolids content of the coating formulation.With water-based emulsions, it is possibleto coat at solids contents of up to 50percent. However, if the goal is toachieve a dry coat weight of about 1 gsm,this puts a severe limit on the choice ofequipment. Condensation chemistrylimits solvent-based coating to somethingunder 10 percent solids; and at this lowconcentration, a wide range of Mayerrods and gravures is available to adjustthe final dry film coat weight to exactlythe desired amount.

8

Figure 4. Typical pan-fed, three-roll differential offset gravure. Figure 5. Alternative Feed System.

Coat weights as low as 0.1 gsm and as highas 3 gsm have all been achieved easily withdirect gravure and coating formulationsranging from 1 to 20 percent solids – thelatter more with addition-curing materialsthan with condensation types. Very lowcoat weights are a little more difficult toachieve with Mayer rods.

Both Mayer rods and gravures do havesome needs of their own. Neither the gapbetween wires on the rod nor the bottom ofthe cells of a gravure empty completely.None of these solvent-based silicones havean infinite bath life, so slowly andinevitably the cells and gaps between wiresbecome plugged with gelled silicone,reducing the volume that can be deliveredeffectively. Hence, both kinds ofequipment need regular cleaning to removegelled silicone. Monitoring either coatweight or coverage (dye check) can give agood indication of equipment cleanliness.Coat weight measurement and coverageassessment will be covered later under“Testing.”

The advantage of the air knife is that nomoving parts touch the substrate, socleanliness is not a problem. This methodis not very satisfactory for solvent-basedcoatings due to solvent loss and viscositycontrol difficulties, but it works very wellwith water-based silicone emulsions wherethe use of viscosity-control agents isrelatively straightforward. The coating of

diluted coatings by the describingmethods is so well established that it doesnot merit further discussion. Coatingwith 100 percent silicone solids, however,is a much more demanding affair.

Although achieving 1 gsm coverage iseasy out of a dispersing medium such assolvent, it is more difficult at 100 percentsolids. A gram of silicone is a ball ordrop about 5 or 6 mm in diameter.Spreading this out over a square meter isa significant challenge and will not beachieved with a brush or finger. Whereasthe same material dispersed in 20 gramsof solvent easily spreads all over, and the19 grams of solvent is allowed toevaporate. Both the flow-outcharacteristics and the accuracy ofcoating become more demanding withoutthe solvent.

The two most common methods usedtoday to apply 100 percent solids coatingsare three- or four-roll differential offsetgravure and the multi-roll smooth rollcoater.

The most demanding aspect of the offsetgravure (OSG) system is the balance ofthe cell size of the gravure and the speedof differential between the gravure andthe applicator or offset roll.

Figure 4 illustrates a typical pan-fed,three-roll differential OSG. Not all nipsare film-splitting nips. This is not always

necessarily so, but it is the most usualconfiguration. Typically, this setupwould use a gravure with 180 to 220 cellsper inch of either a quadrangular type orthe connected quad (CQH) type; othertypes, such as helical or pyramidal, arealso used. With such a cell size, theapplicator roll may turn at 3 to 5 times thespeed of the gravure. The web speedwould then be within 5 percent of theapplicator.

Such a gravure is very fine, easilydamaged and plugs quite frequently. Thetemptation is to use a much coarsergravure of 100 to 120 cells per inch.However, this will necessitate anapplicator roll-to-gravure roll differentialthat is maybe ten times faster. This willproduce a poorer silicone film and a lesssatisfactory coating and, in addition, willput a lot of shear strain on the silicone,which will shorten bath life.

Variations on this system include nip-fed,double doctor feed tube-fed, reverseapplicator, and variations in both gravurematerial and doctor blade angle andmaterial.

In many instances, chrome-plated gravurerollers have given way to laser-etchedceramic-coated rollers. The doctor bladefor the gravure, shown in a negative anglein Figure 4, can be a conventional trailingdoctor blade if the mounting issufficiently strong to

9

Figure 6. Five-roll coater with typical speeds.

resist hydraulic ramping effects on theblade at higher speeds. An alternative feedsystem is shown in Figure 5.

The double doctor blade feed tube systemenjoys the same advantage as the nip feed.That is, less coating is needed to prime theequipment than with a pan-fed system.However, both are prone to air entrapmentproblems, which can be eased by suitablesilicone viscosity selection, by carefulequipment design and by speed of running.

By and large, the offset gravure has beenthe method of choice in the USA. InEurope, the multi-roll smooth roll coaterhas enjoyed more widespread acceptance.This is shown in Figure 6. Note that this isa nip-fed system, and all nips are filmsplits. The rollers alternate rubber/steelthroughout. Typical speed settings areshown, but many variations will workwell. In general, speed differentialsbetween contacting rollers are best kept toless than four to one.

Both methods have their advantages anddisadvantages. The OSG is cheaper,simpler and does all that is expected of it.The multi-roll coater will achieve a lowercoat weight with better coverage and hasno gravure to plug up, but it is complicatedand needs to time to be optimized in termsof nip pressures and roller differentials. Italso costs more.

Both methods have a tendency to be speeddependent; that is, coat weight changes asmachine speed is altered. Both aresusceptible to hardening of the rubberroller as silicone swells the rubber rollsand then cures in the roller. Silicones with“infinite” bath life – electron beam-curlingsilicones, for instance – do not do this tothe rubber rollers.

Both methods work best if the applicatorroll is cut to suit the web width, but thisnecessitates a quick-change system forapplicator rolls and further added expense.The debate over the merits of these twosystems has continued for several years. Ifthe minimum coat weight expected formthe equipment is achievable by OSG, thenboth will do a satisfactory job. There areother solventless silicone coating heads,but none have gained the acceptance of theOSG or multi-roll coaters.

To sum up, for dispersion coatings (eithersolvent- or emulsion-based), several well-established methods exist that will achieveany desired coat weight with goodcoverage on just about any usablesubstrate. For 100 percent solids orsolventless silicone coating, there are twomain methods from which to choose,neither of which will get down to very lowcoat weight with good coverage. Sincefilmic substrates need very little silicone,100 percent solids coating of thesesubstrates

is less economical than desirable. There isstill an unmet need to achieve 0.2 gsmwith excellent film continuity on, forexample, polyolefin films.

THE CURING STAGEThermal curing only will be considered inthis section; ultraviolet and electron-beamradiation curing will be discussed under“Future Developments.” Silicone releasecoatings can be coated out of solvent,water, or 100 percent solids. Immediatelythereafter, they all need to be cured byapplication of heat. Although microwaveswill evaporate water and infrared will dryoff solvent, the most prevalent method, byfar, for curing silicone release coatings ishot forced air. Oven designs come inmany forms. Some common types includearch dryers, which help enormously withweb tension control; and air flotationovens, which avoid contact of the webwith the oven rollers. These ovens applyheat from both sides of the web and arevery efficient heaters. Most new ovens areair flotation types, but some arch orstraight-through ovens with only top-sideheating have been built. This latter type isan attempt to control moisture loss fromcellulosic (paper) substrates.

For dispersion silicones, there are differentcuring requirements for emulsions andsolvents, and both are different fromsolventless requirements.

Consider emulsion drying and curing. Aswill be discussed in the section on“Formulating Release Coatings,”emulsions often contain water-solubleorganic extenders or thickeners as well assilicone release material. Thus, thesequence of events is as follows. First, theuniform, wet film of coating enters theoven, and water begins to evaporate. Thesilicone emulsion particles begin tocoalesce, and the thickener raises theviscosity of the film as water leaves. Thesilicone must coalesce to an oil and moveto the surface before it either begins to gelor is prevented from moving by the film-forming of the organic. Second, when thesilicone

10

Figure 7. Effect of curing temperature on release values. Figure 8. Effect of molecular weight on release force.

oil has formed a silicone-rich surface, therest of the water must evaporate. Finally,the silicone must cure. If the silicone curestoo quickly on the surface (and if theorganic material forms a surface film also),water will be trapped under the film andwill explode holes in it later as added heatforces the water to volatilize.

If too much heat is applied too soon, notonly is the above phenomenon encouraged,but the stratification of the organic andsilicone is hindered, resulting in too muchorganic in the surface and impaired releaseproperties.

Thus is the case for zoned oven heatingmade. Relatively gentle heat is applied atfirst to remove the water slowly and raisetemperatures to coalesce the silicone.Next, somewhat hotter temperatures areused to remove the last of the water andenhance silicone mobility. Finally, hightemperatures are applied to cure thesilicone well and harden the organic.

For solvent-based coatings, the scenario isa little simpler. All that is needed is toremove solvent and, in some cases, volatilecomponents such as catalysts, crosslinkersor additives before they become trappedbeneath the surface skin. Again, somewhatless aggressive heating in the first apart ofthe oven will remove solvent and begincuring. Too much heat too quickly cancause a race between curing and

catalyst evaporation, for example. Withcondensation systems, it is quite easy toevaporate the tin salt catalyst tooprematurely, leading to tin oxide depositsin cooler parts of the oven.

For solventless silicone, the situation is alittle different. With 100 percent solids,the uniformity of the coating as it leavesthe coating head is often poor – a result offilm-splitting effects. Unlike solventcoating, the viscosity of the coating maybe over 300 cp, so it needs time to levelout. This leveling is accelerated bytemperature, but only if curing is nottriggered. Also, the crosslinkers can beevaporated by too much heat too early.So, again, the best situation is a zoned,ramped oven, cooler at the front.

In reality, ovens range from 15 to over300 feet and from single-zone to six oreight zones. Speeds range from 50 to2000 feet per minute. These conflictingcircumstances mean much curing is farfrom optimum. It is not possible to givecuring requirements in a table form. For agiven formulation, the time andtemperature required to achieve anonmigratory, nonsmear, well-anchoredfilm can be defined, but they will beabsolutely individual to each oven.

Clearly, when a coating is still oily orsmeary and transfers to the adhesive orthe backside of the substrate when it iswound up, it is not cured well

enough. However, even the best-curedand aged coating may be put against anadhesive and, on separation, silicone canbe detected on the surface of theadhesive. True, nontransfer of materialfrom one surface to another is probablynot achievable with a soft elastomericfilm such as an unfilled silicone releasecoating. The questions, then become:What is acceptable? How will it bejudged? How is that state of curereached? Unfortunately, the answers areas varied as the applications. A fewexamples: For large labels applied toclean surfaces, quite large amounts ofsilicone transfer can be tolerated. Forvery small, stiff labels applied to curvedsurfaces, very little can be tolerated. Forcarbon-fiber pre-preg laminates, virtuallyno silicone transfer is acceptable.

Nor is silicone transfer the only measureof state of cure. In a polyfunctionalsystem, even after prolonged heating,substantial numbers of unreacted, butreactive, groups still remain, unable tocure because of steric separation from thecorresponding reactive group needed toeffect cure; e.g., vinyl on polymer andSiH on crosslinker. Although to allintents and purposes the film may becured, it is still reactive and not inert.This may lead to chemical reaction withmaterials subsequently brought intocontact with it. This causes release levelinstability.

11

In the food industry, extraction of curedmaterials by water, alcohol, or hexane isoften used to test inertness or cure. Zeroextractable content is another resultachieved asymptotically.

In the release of pressure-sensitiveadhesives, it is often possible to plot curetemperature (for a constant exposure time)and release value. The point where theserelease values become constant could beinterpreted as the minimum requirementfor curing. Unfortunately, this point differswidely for different adhesives. Someexamples are shown in Figure 7.

Points A, B and C could all represent awell-cured coating for the adhesive inquestion. All three may representsignificantly more demanding conditionsthan are necessary for supermarket labels,and all three could be totally inadequate fora food-grade application. Thus, physicalappearance (smear or no smear) mayindicate cure – not too well in absoluteterms, but well enough for, say, asphaltpackaging.

Tape detackification, as a measure ofmigration, is a possibility; but percentextractability is more demanding and morequantitative. However, the question,“What is the acceptable limit,” is pertinent,as zero extractable will be nearlyimpossible to achieve if not to measure.

Release values are fine for a givencircumstance but are a backward-lookingmeasurement since, if the level is too high,the material made is wasted. Releasestability is an excellent measure but, again,occurs too late to make an on-the-spotjudgement. Silicone migration can bedetected on adhesive surfaces by a numberof means, and contact angle of wettingsolutions is a quick, easy test which offersgreat possibilities. However, it needs to berelated to acceptable performance.Crosslink density measurements are a fine,fairly absolute indicator of cur estate, butthe normal test methods are hopelesslycumbersome.

This leaves us in the real world whereeach coater uses one or a combination ofthese methods to make a judgement of thestate of cure of a given siliconeformulation and pronounces it fit or notfor a given application. There is stillroom for much science here!

FACTORS AFFECTINGRELEASE PERFORMANCERelease performance is normallymeasured by the force necessary toremove an adherend from the releasesurface. It is quoted as a force per linearmeasurement, the latter being the width ofthe adherend at right angles to thedirection of removal. This measurementis affected by a very large number ofpotential variables. These will bediscussed under two headings:fundamental and practical. The formerwill include control of the elastic natureof the silicone, surface nature of thesilicone, chemical type of the adhesive,elastic nature of the adhesive and otherfactors inherent in the materials. Thelatter heading will discuss the impact ofdelaminating speed and angle, and theeffects of imperfect coverage of thesubstrate by the silicone. It will revisitthe impact of state of cure, examine theimpact of the thickness of both thesilicone and adhesive layers, and lookbriefly at the impact of temperature andhumidity on release performance.

Fundamental factors inherent in thenature of the materialsOne of the difficulties in writing anaccount such as this is that, viewed today,it ought to seem a coherent, logicaldeductive story. In reality, most of theprogress has been made empirically;much research has gone to explain knownphenomena and has rarely predicted newdevelopments. This is particularly true inthe fundamental work done on the natureof the materials and the impact on releaseperformance.

The earliest solvent-based releasecoatings were based on a silanol end-stopped polydimethylsiloxane of about5000 to 6000 siloxane units in length.Such a polymer is heavily coiled and

not at all a long, thin, gently undulating“string,” although it is often portrayed assuch. When such a polymer iscrosslinked into a silicone elastomer, thelong backbone is only attached to thematrix at each end. This provides a verylarge scope for deformation of theelastomer by displacing and straighteningof the coils in the polymer chains. Thus,careful measurement will show that evena thoroughly cured coating of this naturewill have an elongation at break ofseveral hundred of percent.

Later coatings were developed base donthe same chemistry; the only differencewas the free chain length, or molecularweight, of the silanol end-stoppedpolymers. Polymers of approximately600 siloxane units and even 20 units wereused. These coatings exhibitedsignificantly different releaseperformance – different not only inrelease value at a given peel rate but alsovery different in the relationship betweenrelease force and peel rate. (See Figure8.)

These shorter chain length polymers havesignificantly less extensibility availablethan the long polymers. In practicalterms, the highest-molecular-weightpolymers give soft, very elastic coatings.The 600-unit or so polymers give fairlyhard coatings with just a few percent,maybe 20 or so, elongation at break.. Thevery short chain polymer gives a brittlecoating that is almost inflexible withoutshattering, and elongation at break is toolow to measure. Crosslink densitymeasurements reveal magnitudes ofdifference between all three. From suchempirical beginnings has developed anunderstanding of the rheological impactof both release coatings and adhesives onrelease performance.

Although these condensation systemsoffered quite a wide range of release toaccommodate a wide range of adhesivetypes, there was still a need for releasevalues higher than those available fromsuch systems. Incorporation of siliconeresins was found to raise release valuessatisfactorily, at least at slow peel rates.

12

Figure 9. Effect on release values of adding resin to coating. Figure 10. Comparative high-speed release profiles.

The impact of such resins was not linear,however, and the familiar hockey stickcurve was developed to show the impact ofresin addition. (See Figure 9.)

High-speed peel testing did not becomegenerally available until the mid-1980s.Until that time, most release testing wasdone in the 6 to 600 inch per minute (ipm)range. Converting, printing and matrixremoval in the label industry and slitting,rewind and end-use in the tape industrywere done at speeds much higher than 600ipm. This led to a situation wherematerials might test well but performpoorly in practice and vice versa. The useof resins as high-release additives was agood illustration. Addition of resins at agiven level might increase release resultsquite satisfactorily at low peel rates, but thefinished laminate might show noimprovement at all in convertibility.Indeed, sometimes convertibility actuallydeteriorated. Given the advantage oftoday’s high-speed peel testers we can nowsee why. (See Figure 10.)

The increase in release force at low peelrate is clear and unequivocal, but at highspeeds, release is actually lower. It wasmany years before an understanding of thisdifference was forthcoming.

In the emulsion area, it was not possible toemulsify the highest-molecular-weightpolymers, so only products based onlower- molecular-weight polymers were

available. This did limit the performancerange available to emulsion users. Inaddition, no satisfactory emulsion of theresin-based additives was offered.

With some additive chemistry alreadymentioned and some alternative tin-basedcatalysts, the condensation systems hadreached the limit of their development. Arange of different-molecular-weightsilanol end-stopped polymers; a resin-based high-release additive; three or fourorganotin salt catalysts offering differingsolubilities in various solvents, differingvolatilities and differing cure rates; somefast-cure additives and an anchorageadditive or two constituted the availableoptions. This was true as much as 15years ago and remains true today.

As environmental pressure againstsolvent emissions grew, the limitations ofemulsion technology became moreapparent, and attention turned to 100percent solids or solventless systems.Condensation chemistry proved unable tomeet the challenge, and additionchemistry was adopted. As stated earlier,this chemistry made it possible to makemultifunctional polymers, not just end-stopped ones:

This had great implications for release

control and cure rate improvement.

Initial work was concentrated on asolventless system. Condensing twentyyears of development, both in releasetesting and materials, the situation todaycan be summarized as follows:

Polymers can be made with only terminalreactive groups, only side chain reactivegroups or both:

For a solventless system to be coatable onthe roll coaters described earlier, it isdesirable that the polymers should be nolonger than approximately 300 siloxaneunits, and preferably in the 100- to 200-unit chain length range.

The release behavior of these polymers isillustrated in Figure 11.

These differences, however, are relativelysmall compared with the range available incondensation solvent-based systems. (SeeFigure 12.)

These graphs change significantly as theadhesive mass is changed, but therelationships remain reasonably constant.As can be seen in Figure 11, the terminal-functional solventless polymer “A” gives

13

Figure 11. Impact of polymer morphology on release profile. Figure 12. Comparison of solvent-based and solventless polymers.

somewhat more elasticity than themultifunctional polymer “C,” but both givelow release values and flat release valueprofiles with peel rate. After 30 years or soof condensation-based systems, theindustry was attuned to a certain type ofproduct; and the move to solventless didnot fit well, as release values weregenerally too easy. The use of resin-based,high-release additives ceased to be asolution for special needs and became anessential component of most constructions.Unfortunately, the same tendencies persisteven today. These resin-based materialswork well at low peel rates but not so wellat high rates.

Fortunately, help was at hand. A thesame time silicone release coatings weremoving away from solvent, the samemove was afoot in adhesives. Until themid-1970s, solvent-based rubberadhesives were at least 80 percent of thelabel and tape industries. It provedimpossible to either emulsify thesematerials or make a solventless version.Consequently, major changes inchemistry and materials were required.Two contenders emerged and stillcompete neck and neck today: water-based acrylic emulsions and rubber-basedhot-melt adhesives. These have verydifferent peel rate profiles than solvent-based rubber adhesives. It has been

possible to marry the emergence ofsolventless silicone to water-based andhot melt adhesives and make productsthat satisfy market needs.

Solvent-based SBR adhesives, forexample, are high-molecular-weightmaterials, most of which do not requirecuring, just solvent removal. The dryadhesive has both high physical strengthand great elasticity. This shows in a peelrate/adhesive strength curve and also in arelease force/peel rate curve. (See Figures13 and 14.)

The water-based acrylic has greaterelasticity but lower strength, and the

Figure 13. Solvent-based SBR adhesive – peel strength. Figure 14. Solvent-based SBR adhesive – release force.

14

Figure 15. Typical release profiles for various adhesives. Figure 16. Release profiles with water-based acrylic adhesive.

hot-melt, vice versa. This yields therelease force/peel rate comparisons shownin Figure 15. These are typical examplesof adhesives released from a solventlessrelease coating. Clearly, the need for high-release materials is greater for use withhot-melt adhesives tan with water-basedacrylic adhesives. This is true in practiceas well as in theory.

In Figure 13 we can see the peel strengthdrop at a certain peel rate. This is normallyexplained in terms of Tg phenomena. Tgis usually thought of in terms of thetransition from a glass-like state to arubbery state as the temperature is raised.However,

one can also encounter the same apparentstate change when frequency is usedinstead of temperature. Most people arefamiliar with breaking thread bysnatching instead of a steady pull. Themolecules do not have time to relax andmove around each other, so they breakinstead. A similar phenomenon occurs inadhesives, often accompanied by asignificant amount of cohesive failure inaddition to adhesive failure.

This is not a treatise on adhesives orphysics, but some discussion is necessarysince precisely the same phenomenon isoccurring in the silicone release coatingas peel rates are

increased. The change from rubberybehavior to glass-like behavior inadhesives is very often marked, in thepractical sense, by a change from smooth,quiet release to raspy, noisy release. Thisphenomenon is particularly common inhot-melt adhesives, which just happen tohave a high Tg! It is not nearly somarked in water-based emulsions,particularly acrylics, and is pretty much ararity in a pure polymer-based releasecoating. However, the resins used ashigh-release additives are typically solids,even at high temperature, and areparticularly hard and glass-like.Incorporating these resins into a releasecoating

Figure 17. Impact of resin on release profile of a hot-melt adhesive. Figure 18. Water-based acrylic release test.

15

Figure 19. Hot-melt adhesive release test. Figure 20. Effect of resin on storage modulus of lightly crosslinkedpolymer systems.

effectively raises the Tg of the curedcoating. Thus, even water-based acrylicadhesives may show the onset of raspyrelease – not exhibited in the absence ofthe resin additive – at elevated peel speeds.Hot-melt adhesives will, therefore, presenta major problem with a resin-containingrelease coating if raspy, noisy release is tobe avoided. Figure 16 and 17 illustrate.

Continuous monitoring of a release forcemeasurement can be even more revealing.Figure 18 shows a water-based acrylicreleasing from a solventless releasecoating with no resin high-release additive.

Figure 19 shows a hot-melt adhesivereleasing from a coating with about 25percent resin. Both are at 10 meters perminute peel rate.

Figure 18 represents smooth, quiet release,and Figure 19 raspy, noisy release. Thenoisy release is associated with cohesivefracture of both the silicone coating andthe adhesive. It is due to the more glass-like behavior of the materials at thisfrequency than that of the lower Tg (andhence more rubbery) materials in Figure18.

The question remains, “Why do theseresins raise release at low peel rates?”

Figure 21. A typical release profile.

We have seen why resin effectiveness isgreatly reduced at higher peel rates. Theanswer to our question is not absolutelyclear, but it dos seem that the rubberynature of the crosslinked polymer isaltered by the resin, which acts as a smallparticle filler. The rubbery plateau of thecoating is moved to another domain, andthis may be the major effect. Certainly,the cohesive strength of the silicone film isincreased by resin incorporation. Figure20 shows the change in “rubbery plateau.”

Returning now to the addition-curedpolymers: clearly, the multifunctionalpolymer will produce a higher crosslinkdensity than the terminal-only polymer.This will alter the basic nature of anunmodified coating somewhat, themultifunctional polymer being more glass-like than the end-functional one. Thisdifference will then accentuate the resintendencies.

However, release values, profiles and theircontrol are not the only factors to beconsidered when formulating releasecoatings. These other factors, which willsometimes be in conflict with pure releaseconsiderations, will be examined under theheading “Formulating Release Coatings.”

16

Figure 22. Figure 23. Figure 24.

The practical control of releaseperformanceTo obtain some kind of benchmark,consider one coating type for a while. Itmakes no great difference whether it is anemulsion, solvent-based or solventlesstype. Some factors affecting release arethe same for all three.

The relationship between release forcesand peel rate has already been mentioned.Figure 21 (page 15) can be taken astypical.

Clearly, measurement at 12 ipm (0.3meters per minute [mpm] ), 400 ipm (10mpm) or 12,000 ipm (300 mpm) will givevery different release values.

Just how different will depend upon theadhesive above all, the silicone and someother factors, too.

The most important of these other factorsis the angle at which the adherend ispulled from the release surface. If apressure-sensitive laminate is pulled apartin a free-hanging unsupported state(Figure 22), the angle between the twosubstrates will reach a value thatminimizes the force applied.

In matrix removal operations on the press,this angle is often controlled to givesmooth removal. In release testing, theangle is fixed, either at 90

or 180 degrees. At least that is the intent.In reality, due to the stiffness of bothsubstrates, the angle is significantly lessthan 90 or 180 degrees. (See Figures 23and 24.)

So, substrate basis weight, andsubsequent stiffness, is a major variable.Furthermore, in reality on the press, bothlabel stock and liner are pulled, whereasin testing, either one of the other ispulled. A major influence on stiffness ofpaper substrates, which are by far themajor proportion of the market, ishumidity. Figure 25 shows the impact ofhumidity on release force, and Figure 26shows the impact on release profile.

Figure 25. Effect of humidity on release values. Figure 26. Effect of humidity on release profile of an emulsion acrylicadhesive-based laminate made with paper backing and facing.

17

Figure 27. Figure 28. Effect of film thickness on release.

observed. The greater the thickness ofeither, the greater their energy-absorbingcapabilities as they stretch and, thus, thehigher the release forces required.However, for stiff, fairly inelasticexamples, the impact may be very small.Figure 28 shows the impact of coat weight(or layer thickness) on release values forsome silicones. “A” represents the amountnecessary for total coverage – less allowspaper fibers to impact release values.Figure 29 shows how adhesive thicknessaffects release force.

Now it is time to re-examine state of cure.If a coating is “fully cured,” whatever thatmeans, it should reach a

state of constant physical parameters interms of hardness, elasticity, strength, etc.,and hopefully, inertness. Releaseperformance should also be consistent andrepeatable, and release values should bestable. This state is easier to achieve withsome polymer types than others. Also,some adhesives can tolerate significantundercure, while others cannot. Sincesuch a taste of “total” cure is achieved atthe cost of long exposure to curingstimulation, be it thermal or radiation, thesubstrate itself may suffer significantly interms of impoverished physical properties.Thus, such a state of cure is not alwaysdesirable. In reality, very few

Temperature will affect the rheologicalstate of the adhesive, in particular, andalso the state of the silicone and substrates.Clearly, for release testing, temperatureand humidity are major variables inaddition to peel rate and peel angle.

Although the effect of these variables hasbeen discussed from a release testingstandpoint, they do have the sameinfluence in practical use. For example,asphalt is easily removed from releaseliner when cold and hard, but removalbecomes increasingly difficult as theasphalt becomes more rubbery. A furtherexample is the conversion of a givenlaminate for different size labels. Largelabels usually convert and matrix stripeasily; but the same material, when used tomake very small labels may prove toostiff, and the labels will fit off as thelaminate moves over small-diameterrollers. (See Figure 27.)

Another major factor affecting release isthe degree of silicone coverage of the baseliner. If greater or larger portions of thebase are left either uncoated orincompletely coated, the adhesive willcontact the base, and release values will beboth higher and less stable with time.Coverage assessment will be examinedunder “Testing.”

Adhesive and silicone coating thicknessalso affect release force measurement.These effects are simply

Figure 29. Effect of relative thickness and adhesive type.

18

release coatings are cured to such a state,and all are undercured to some degree orother.

What, then, is the impact of thisundercure? The answer is totally adhesivedependent. Most rubber-based adhesives,solvent or hot melt, are relatively inert tounreacted reactive groups in the silicone.On the other hand, crosslinkable solventacrylic adhesives can and do react stronglywith SiH functionality in the silicone. Thisphenomenon is known as acrylic lock-up.Most water-based adhesives, either acrylicor rubber-based, exhibit similar tendencybut to a lesser degree. With undercuredsilicone, release forces may be a littlehigher with even inert adhesives, simplybecause the elasticity of the silicone ishigher due to a lower crosslink density or,easier said, a softer silicone.

For any given construction, there is aminimum degree of cure that will producean acceptable product. This degree of cureis usually the point most users achieve, andany greater curing effort is perceived as athreat to the substrate. Release stabilitywill always be improved with improvedcuring, and release values will usually belower. Other properties such as adhesivetack, subsequent adhesive strength andextractable content of the silicone film willall improve with better curing.

In summary, the major factors affectingrelease were brought together under twoheadings. The basic chemistry, coatingmethods, and the decisive influence ofcuring were reviewed. Coatingformulation and the building blocks usedwill now be discussed.

FORMULATING RELEASECOATINGS

Condensation typeBasic polymers:

Where n ranges from ~20 to ~6000.

Crosslinkers:

Where n is ~40.

Catalysts:

Where OAc can be laurate, hexanoate,acetate, etc.

Fast-cure additives:

Si(OR)4

Where R is of a type such asCH2CH2OCH3, etc.

Anchorage additives:Epoxy-functional materials.

High-release additives:Resin-based of the so-called M:Q resintype.

The selections made will reflect thedesired release performance. The basicpolymers range from low-viscosity fluids,where n=~20, to thick gums, wheren=~6000. While permitting release forcevariation, the problem of solventdispersion soak-in becomes more severeas the polymer molecular weight isreduced. Blends of these polymers areusually used whenever the 20-unitpolymer is involved. Use of low-molecular-weight polymers is bettersuited to substrates with excellent holdoutsuch as polycoated kraft or films. Curerates differ somewhat, but all will cure atroom temperature after solvent removal.Catalyst choice will be a reflection of

solvent choice, oven temperatures andregulatory status. Anchorage additivesare occasionally necessary on somesubstrates. They do not function onpolyolefinic surfaces.

In the emulsion field, the lower-molecular-weight polymers can bemechanically emulsified as can the morehydrolytically stable catalysts. Emulsionsystems do not have fast-cure additives orgood high-release additives, but oftenthey do not need them; and they do enjoythe ability to be used with water-solubleorganics such as polyvinyl alcohol,carboxymethyl cellulose, starch and ahost of others. They can also be usedtogether with other organic emulsionssuch as polyethylene emulsion. Thisopens up significant opportunities notopen to solvent-based materials.

Addition cured materialsBasic polymers:Both terminal and multifunctional vinylpolymers.

Where n+m ranges from 100 to 5000, andn ranges from 96 to 100 mole percent(i.e., no side chain – terminal only).

Normally, in solvent products, high-molecular-weight gums with side-chainvinyl groups are used. Solventless andemulsion products use much lower-molecular-weight materials and will useboth terminal-only and multifunctionalpolymers. The influence of these onrelease has been discussed, but there areseveral other implications, also. In apolymer with five or six reactive groups,only one need react to tie the polymerinto the gel. If the polymer has only tworeactive groups, one must still react. So,in one case, one in five or six must react;in the other, one in two. This reflectsitself in several ways, but the mostobvious is in percent extractable materialat various states of cure. Both kinds ofpolymer can be cured to very lowextractables, but for any given set of cureconditions, the multifunctional polymerwill

19

Figure 30. Differential scanning calorimeter trace.

show less extractables. This is anindication of cure rate. However, otherindicators will show that the terminal-functional-only polymers reachcompletion faster – stability of releasewith a solvent crosslinkable acrylicadhesive, for instance, or differentialscanning calorimetry (DSC). (See Figure30.)

So, as in most aspects of release coatings,things are different, not necessarily betteror worse. The use of a multifunctional orterminal-functional polymer as the basisfor a solventless or emulsion addition-curing coating is very much a question ofsuitability to specific needs. Apart fromfunctional issues such as curing andrelease properties, other issues such asworking life, coatability and cost arerelevant to the polymer section. Today,with a range of inhibitors, bath life is onlya problem under special circumstances(one of which is gravure clogging), butcoatability merits a few words.Differences between various polymers canbe seen in coverage efficiency, levelingspeed (removal of film splitting lines),gravure cell filling and emptying, and inthe development of a fine mist or sprayaround the final transfer nip where siliconeis actually laid onto the substrate. Thesedifferences are often as much a function ofpolymer molecular weight as of polymerarchitecture, but multi- and terminal-functional polymers do exhibit differences.

Crosslinkers:Both homo- and co-polymers of SiHfunctionality are used.

Homopolymer “A” is chosen for rapidgelation, good anchorage to substrates andgood cohesive strength in cured film. nranges from 5 to 60.

Copolymer “B” is chosen for good bathlife and good final cure characteristics.n+m ranges from 5 to 60, and m from 50to 90 mole percent.

Blends of “A” and “B” types are oftenused to blend properties. In solvent-basedmaterials, “B” exhibits poorer anchorageproperties than “A” and is not widely used.

Catalysts:Noble metal organo complexes are used.To improve solubility in 100 percentsilicone systems, these are usuallyplatinum or rhodium organosiliconecomplexes. In solvent-based products,nonsilicone complexes

can be used. The normal level is 50 to 150ppm platinum metal, based on totalsilicone content.

Inhibitors:These were listed under the chemistrysection, but suffice it to say, someinhibitors can be packaged together withthe catalyst with no damaging effects.Others, such as acetylenic alcohols,cannot. Inhibitors usually range formstrong inhibitors with long, stable controlon the bath life, but higher initiationtemperatures, to low-initiation-temperatureinhibitors with barely adequate low-temperature inhibiting properties.Solubility in 100 percent silicone systemsis always a concern, and often low-temperature storage of formulatedproducts can cause the inhibitors to comeout of solution.

In solvent-based systems, a wider rangeand greater concentrations are possible, sooften spectacular curing performance canbe coupled with good working bath life.Also, inhibitors can be blended to combineproperties.

In emulsions, the separation of catalyst,polymer and crosslinker is easily achievedvia separate oil-phase particles. Additioncuring emulsions offer some veryinteresting capabilities, usually consistingof the same basic materials as solventlesssilicones.

High-release additives:These are essentially the same kind ofresinous materials used in condensationsystems, but with the added advantage ofbeing functionalized. Since they aresolids, soluble in hydrocarbon solvents,their use in solvent-based materials isstraightforward. For emulsions andsolventless materials, however, they mustbe dispersed in a silicone polymer. Thisautomatically limits the amount of resinthat can be incorporated into a coating,since the polymer is added simultaneously.If the resin content is too high insolventless systems, coating problemsarise as the viscosity rises rapidly withresin content.

20

Figure 31. Effect of solvent choice on bath viscosity with aging. Figure 32. Differential scanning calorimeter (DSC) trace.

Studies of resin molecular weight, type andfunctionality have shown some impact onthe efficiency of raising release, but thesignificance is minor compared with theeffect of resin concentration. The efficacyof high-release additives is a major concernin both emulsion and solventless areas.

In concentrating on solventless silicones, Ihave reflected the current state of themarket, but solvent-based systems do meritfurther consideration. In bothcondensation and addition-curing systems,the choice of solvent used for coating is acritical factor. The solvent controls theoverall density of the coating solution.This needs remembering since, while thecoating mix is most often made up byweight all coating methods lay down avolume.

The spreading or wetting of the coatingsolution on various substrates is a criticalfactor in achieving both good coverage andanchorage. Solvents with a surface tensionbelow that of silicone (21 to 22 dynes percm) will spread and wet well on mostsubstrates. Thus, hexane and heptane aregood wetting solvents and will coat almolecular-weight silicone polymers well.Toluene will not, and anchorage is often aproblem; on low-energy substrates such aspolyolefins, beading up or reticulation is amajor problem.

Use of low-molecular-weight siliconepolymers exacerbates this problem.

Mixtures of solvents can be used toadvantage, as heptane is sometimes alittle too volatile for use alone.

In addition to wetting and spreadingdifficulties, bath life can be very muchaffected by solvent choice. This is trueboth for condensation and addition types.

Figure 31 illustrates rate of increase inviscosity for an addition-cured siliconecoating system in various solvents.Usually, there is a practical viscosity limitfor gravure or Mayer rod coating. Thus,solvent choice greatly influences bathlife.

The use of solventless systems out ofsolvent is quite common. Thedisadvantages are that such low-molecular-weight materials give verylow-viscosity coating solutions, andpenetration into the substrates is a majorproblem. Also, such low viscosities canmake the actual coating process a reallymessy operation. However, one can laydown very thing coatings of solventlesssilicones, which is desirable for filmcoating, for example, but solvent choiceis critical.

Some solvents lend themselves muchbetter to solvent recovery than others.Some solvents are nonflammable, and areattractive as such. However, if anoptimum job of siliconizing is to bedone, these kinds of considerations must

come second to the effectiveness of thesolvent as a coating, curing and bathmedium.

TESTINGThis subject is easily divided into twoapproaches: testing of the raw materialsbefore use and testing of finishedproducts.

Raw materialsOnly silicone release coatings will beaddressed, not substrates and adhesives.

Silicones can be measured physically;e.g., viscosity, specific gravity,nonvolatile content, percent vinyl,hydride, silanol, alkoxide, etc. These areimportant tests related to consistency;they tell us nothing about functionalstability. The two major areas offunctional suitability are ability to cureand release performance.

Cure rate can be examined by coatingonto a chosen substrate and curing in anoven, then measuring cure by one of themethods already discussed. This is a verysubjective test, but can be quite useful.More recent tests that are much moreobjective are becoming widespread.However, the relationship between thetest and real use is very much debatable.

Differential scanning calorimetry (DSC)and dynamic mechanical

21

Figure 33. Development of cohesive properties with time (curing). Figure 34. Comparative DMA of terminal and multifunctionalpolymers.

content and loop tack. No doubt, time willimprove the usefulness of the informationprovided by DSC.

DMA is a little easier to envision. Aliquid sample ends up as a lump of curedelastomer. Progress is monitored alongthe way. A number of devices exist thatcan do this with varying sophistication andcost. A typical trace is illustrated inFigure 33.

It is useful to look at two systems here –the terminal-functional polymer and themultifunctional polymer. (See Figure 34.)

This is an intriguing difference and

mirrors cure data provided by acrylicadhesive compatibility, for instance.Recent work done with the vibratingneedle curemeter1 (VNC – a simple formof DMA) has begun to show therelationship between rheologicalproperties and release properties, and tooffer hope of release performanceprediction before coating and curing. TheVNC offers information on both cure rateand release value and so is of greatinterest.

Figure 35 shows a series of VNC curves.Curve “A” represents a system that iscuring into a tightly crosslinked gel. Sucha system uses either a very

analysis (DMA) are both finding support.The first measures the heat given out by anexothermic curing reaction. The secondfollows the development of mechanicalstrength as a system cures or hardens. Atypical trace is shown in Figure 32.

• t1 is the onset time (temperature).• t2 is the time (temperature) at which

the maximum heat flow is reached.• t3 is the time (temperature) at which

cure is complete.• “H,” the area under the curve, is the

total heat given out during curing.

Figure 30 (page 19) showed the differenceexpected between a highly functionalpolymer and a terminal-only functionalpolymer. The DSC is a wonderful tool forexamining consistency, but therelationship of any of these parameters tofunctional suitability of a release coating isdifficult to see. Certainly, a smooth, rapiddecline from t2 to t3 is indicative of a“finished” cure. However, since a largenumber of very commercially successfulcoatings do not show this, a littleuncertainty exists. Further, some coatingsare seen to have higher initial, peak andfinish temperatures and yet clearly showfaster cure on a machine in terms ofrelease stability, extractable

Figure 35. Development of adhesive properties with time (curing).

22

short chain terminal-functional polymer, ora multifunctional polymer with severalreactive groups along the chain – two orthree per hundred siloxane units. Such asystem is hard and brittle when cured andwill give very flat release force versusdelaminating speed profiles. Curve “C”represents a system that is soft when cured.Such a system uses a terminal-functionalpolymer, one with no side chainfunctionality. This type of cured filmoffers rising profiles of release force andspeed and will tend to give smooth, quietrelease with hard adhesives. The releasecharacteristics are predictable from thefinal resonating frequency of the gel. Thecure rate prediction comes from the rate ofrise from the liquid state to the cured gelstate. In this case, it is clear that ”A” risesmore rapidly than “C.” In reality,multifunctional polymers do gel morequickly than terminal-functional-onlypolymers. Thus, the VNC curves enable usto predict, at least to some extent, the curerate and release values of the two polymersystems. Curve “B” represents andintermediate state between “A” and “C.”Such a system could come from either alow-viscosity terminal polymer or apolymer of reasonably long chain lengthwith just one or two reactive groups perhundred siloxane units along the chain.

A very similar set of curves can beproduced by the incorporation of high-release additives into any given polymersystem. Since the resins are generallysolids, they are seen by the VNC as veryhighly crosslinked species, and evenmodest levels of resin will greatly increasethe hardness of even the softest curedpolymer films. The impact on releaseprofiles has been discussed earlier. TheVNC offers a somewhat predictive insightinto the use of high-release additives invarious polymers.

Neither the DSC nor the DMA methodshave gained wide acceptance for testingyet, but they show great promise. Neitheris suitable for solvent-based or emulsionsystems.

Release testing before use is clearly acontentious subject, since substrate,

curing conditions, adhesive and testconditions will all need to be tightlycontrolled if such testing is to bemeaningful. Mostly, the preference is toexamine other parameters and trust thatrelease will be in specification.

Release testing of the finished product,laminate, tape, etc., is another matter andis a fundamental activity.

Finished product testingFirst, consider silicone-coated liner.Tests of relevance include coverage, coatweight and state of cure.

Coverage is easily tested on papersubstrates by one of a number of dye stainmethods. Essentially, such a stain colorsthe paper and not the silicone. Non-covered areas are easily seen. However,just how complete coverage needs to beto provide acceptable performance issomewhat more subjective. It is notdifficult, however, to provide a dye stain“standards” match to releaseperformance.

Coat weight measurement is very easytoday, thanks to the emergence ofbenchtop x-ray fluorescence units thataccurately measure silicon content. Toturn this coat weight, however, knownstandard coat weights are needed, so theinstrument is only as good as itsstandardization. Nevertheless, theseinstruments have wrought a mini-revolution in the silicone release coatingindustry. Other methods to exist based onother forms of detection, but are notwidespread. Further methods based on ausage can also act as a broad check.

In addition to being used for releasingadhesive masses, release liners are usedof roil and water repellency, release offoodstuffs, industrial resins, sealant andcaulk masses, asphalt, synthetic leatherand clothing materials, and as barriers toliquids. Testing for such applications isas varied as the applications themselvesand cannot be covered here.

SYSTEM CHOICEThe choice among solvent-based,

emulsion or solventless systems is verymuch dependent on the needs of theoccasion, both in terms of processing andproduct performance. (See Figure 36.) Intoday’s market, more than 65 percent ofsiliconizing is done with solventless. Tenyears ago, less than 15 percent wassolventless.

Looking at the various market segments,the label industry today is dominated bysolventless siliconizing. The tape industrydraws on all three, but solvent-basedcoating does offer some advantages interms of smooth rewinding. In the non-pressure-sensitive sector, the food andpackaging segment uses mostly emulsion-coated silicone, and solventless has asurprisingly small share. What does thefuture hold?

FUTURE DEVELOPMENTSToday, little or not development is beingdone on condensation-curing systems. Wemay be nearing the limits of thermal-curing addition systems, as well, althoughnew catalysts and inhibitors are still verymuch sought after. Faster-curing systemsthat yield more inert cured films are thegoal. Emulsion systems have some way togo to catch up to solventless, soimprovements are likely there; and thedream of an all-water-based facility is stillreal. At present, such facilities have lowerproductivity than solventless/hot-melt set-ups.

Radiation curing has been around for awhile now and has generally proveddisappointing. Why? In non-releasefields, radiation curing produces beautiful,glossy coatings with millisecond cures. Inthe release field, needs for both a highdimethyl content and a reasonable degreeof elasticity have produced releasecoatings with raspy, noisy release, muchslower cure rate than their organiccounterparts, and some release stabilityproblems. Eventually, all these issues willbe resolved, but not in the short term.Perhaps, significant changes in adhesiveswill rescue this kind of hard, inelasticcoating, as happened for thermalsolventless. True UV curing

23

Figure 36. Advantages/disadvantages of solvent-based, emulsion and solventless systems.

Advantages DisadvantagesSolvent-based:

• Very accurate coating control, both in terms of the coverage and coat weight

• Wide range of products, both addition and condensation curing

• Fast, low-temperature cure capability

• Can be used for a very wide range of substrates

• Good release profile range

• EPA and flammability problems; need either solvent recovery or incineration (both costly)

• Difficulties with release control for tight release

Emulsion:

• Water-based (nonflammable)

• Wide range of organic materials available with which to blend (both soluble and latex)

• Excellent coating control

• Coatable by a wide range of methods

• Water-based; damages paper

• Difficult to coat plastic substrates

• More difficult to use on heat- sensitive substrates

• Materials susceptible to freeze-thaw damage

• Drying and curing more demanding than other systems

• Tight release control difficult

Solventless:

• No EPA problems

• No solvent costs or concerns

• Current technology

• No flammability problems

• Curing cycle is the easiest of the three systems

• Coating difficulties - Special coating head required - Get higher coat weights - Coverage concerns

• Tight release control is most difficult of the three

• Need higher base paper quality

has not yet arrived. We do have UV-initiated epoxy ring opening curing, butcatalysts continue to be a challenge. EBcuring faltered on the dual problems ofequipment costs and inerting costs.

An EB system that did not require inertingwould be a big step forward. Currentchemistry does not hold out muchpromise; but there is much

activity in other, nonsilicone fields, andsome may be transferable to releasecoatings. Certainly, the completeabsence of post-cure inherent in EBtreatment is a big attraction.

Finally, in all of these systems, the mostpressing needs is still better releasecontrol. The move to solventless broughtvery easy release and flat release force/

peel rate profiles, effectively lowering thebottom end of the release-force spectrumbut, at the same time, putting the high endfurther away. Release-force elevation withresins is still the only half-way decentmethod, and it is not at all satisfactory inmeeting all market needs. The need is for asystem that is equally effective at all peelrates, which will reach out to the 20 to 30times base level instead of today’s 10 to 15(if we are lucky). We need a system thatworks well with all adhesives and does notimpair anchorage or cure rate, or produceraspy, noisy release.

So, as we look to the future in conventionalcoatings, improved release control might bethe most desirable single development.Improved radiation curing systems andimproved coating methods may also appear.

What of nonconventional developments?Several come to mind:

• Plastic release liners made by extrusionof, say, polyolefin and silicone – nocoating involved

• Non silicone release coatings (difficultfor me to imagine, but certainlypossible, especially with eutecticmixtures of various polyolefins)designed to lower the Tg of the coating

Linerless labels have made an appearanceand an exit. Now, further examples oflabel/label constructions are appearing.Both involve silicone release coatings, andboth have niche possibilities. Both neglectthe fact that the release liner does more thanjust release the adhesive.

What about plastic replacing paper?Certainly, on the label side, significant shifthas occurred. On the liner side, however,paper has made more progress in calipercontrol, flexibility and strength andenvironmental friendliness in recent yearsthan plastic films have. The combination ofcost and low stretch makes paper tough tobeat in the roll label field.

One thing is likely – changes will beevolutionary, not revolutionary, just as theyhave been in the past.

24