5Thermoplastic Elastomers:Fundamentals and Applications
Tonson Abraham and Colleen McMahanAdvanced Elastomer Systems, L.P., Akron, Ohio, U.S.A.
I. INTRODUCTION
In the fifteenth century, Christopher Columbus witnessed South Americansplaying a game centered around a bounceable ‘‘solid’’ mass that wasproduced from the exudate of a tree they called ‘‘weeping wood’’ (1). Thismaterial was first scientifically described by C.-M. de la Condamine andFranc�ois Fressneau of France following an expedition to South America in1736 (2). The English chemist Joseph Priestley gave the name ‘‘rubber’’ to thematerial obtained by processing the sap from Hevea brasiliensis, a tallhardwood tree (angiosperm) originating in Brazil, when he found that itcould be used to rub out pencil marks (2). A rubber is a ‘‘solid’’material thatcan readily be deformed at room temperature and that upon release of thedeforming force will rapidly revert to its original dimensions.
Rubber products were plagued by the tendency to soften in the summerand turn sticky when exposed to solvents. This problem associated withnatural rubber was overcome by Charles Goodyear in the 1840s by subjectingthe rubber to a vulcanization (after Vulcanus, the Roman god of fire) process.Natural rubber was vulcanized by heating it with sulfur and ‘‘white lead’’(lead monoxide) (2). In May 1920 the German chemist Hermann Staudingerpublished a paper that demonstrated that natural rubber was composed of achain of isoprene units, that is, a polymer (from the Greek poly, many, andmer, part) of isoprene (3). In vulcanization the rubber macromolecules arechemically bonded to one another (‘‘cross-linked’’ in a thermosetting process)to form a three-dimensional network composing a giant molecule of infinite
molecular weight. At present the word ‘‘rubber’’ is associated with macro-molecules that exhibit glass transition below room temperature and have‘‘long-chain,’’ ‘‘organic,’’ carbon-based backbones or ‘‘inorganic’’ back-bones typified by polysiloxanes and polyphosphazenes.
‘‘Elastomer’’ is always used in reference to a cross-linked rubber that iselastic (Greek elastikos, beaten out, extensible). An elastomer is highlyextensible and reverts rapidly to its original shape after release of thedeforming force. Entropic forces best describe rubber elasticity (4). However,it should be noted that under relatively much smaller deformation, plasticmaterials and even metals can exhibit elasticity due to enthalpic factors (4).Gases and liquids also exhibit elastic properties due to reversible volumechanges as a result of pressure and/or heat (4). Nevertheless, the term‘‘elastomer’’ is always used in reference to rubber elasticity.
A plastic material is one that can be molded (Greek plastikos), and athermoplastic can be molded by the application of heat. A rubber compound(a blend of rubber, process oil, filler, cross-linking chemicals, etc.) is thermo-plastic and is ‘‘set’’ after several minutes in a hot mold, with loss ofthermoplasticity. A thermoplastic material can be molded in a matter ofseconds, and the molded part can be reprocessed. The viscous character of thethermoplastic melt readily allows control of the appearance of the surface offinished goods. In comparison, the effect of ‘‘melt elasticity’’ of a rubbercompound on end product surface appearance is not as readily controlled.
The origin of the first thermoplastic material can be traced to ChristianSchonbein, a Swiss scientist who broke a beaker containing amixture of nitricand sulfuric acid and used his wife’s cotton apron to clean up the spillage!Unfortunately for his wife, but fortunately for science, he left the washedapron near a fireplace to dry. The cotton apron soon combusted withoutleaving any residue! Schonbein realized that the cotton of the apron wasconverted to ‘‘gun cotton,’’ a nitro derivative of the naturally occurring poly-mer cellulose (1). This learningmay have been instrumental in the preparationof the first plastic by the English chemist and inventor Alexander Parkes in1862. First called Parkesine, it was later renamed Xylonite. This substancewas nitrocellulose softened by vegetable oils and a little camphor. During thistime, elephant tusks, which were used to make ivory billiard balls, amongother things, became scarce. In 1869, motivated by the need to find a suitablesubstitute for ivory, John W. Hyatt in the United States recognized the vitalplasticizing effect of camphor on nitrocellulose and developed a product thatcould be molded by heat. He named this product obtained from cellulose‘‘Celluloid’’ (Greek oid, resembling). Though primarily regarded as a substi-tute for ivory and tortoiseshell, Celluloid, despite its flammability, foundsubstantial early use in carriage and automobile windshields and motionpicture film (3).
A thermoplastic elastomer (TPE) is generally considered a bimicrophasicmaterial that exhibits rubber elasticity over a specified service temperaturerange but at elevated temperature can be processed as a thermoplastic(because of the thermoreversible physical cross-links present in the material).It offers the processing advantages of a highly viscous melt behavior and ashort product cycle time in manufacturing due to rapid melt hardening oncooling.
B. Classification of Commercially Available ThermoplasticElastomers
The TPE products of commerce listed in Table 1 are classified in Table 2 onthe basis of their polymer microstructure. Representative examples areincluded for each polymer class. Segmented block copolymers, triblockcopolymers, and thermoplastic vulcanizates represent a significant portionof the TPE family.
The fundamental aspects of structure–property relationships in ther-moplastic polyurethanes (TPUs), styrenic block copolymers (SBCs) [withemphasis on styrene/ethylene-1-butene/styrene (SEBS) copolymers andSEBS compounds], and thermoplastic vulcanizates (TPVs) produced frompolypropylene and ethylene/propylene/diene monomer (EPDM) rubber wereselected for review in this chapter, as representative of the most commerciallysignificant and the closest in performance to thermoset elastomers.
Table 1 Thermoplastic Elastomer Products of Commerce
Product
First commercialized
(year, company)
Plasticized poly(vinyl chloride) 1935, B. F. GoodrichThermoplastic polyurethane 1943, Dynamit AGPVC/NBR blends 1947, B. F. Goodrich
Thermoplastic vulcanizates possess sufficient elastic recovery to chal-lenge thermoset rubber in many applications, and insights into TPE elasticrecovery and processability are presented based upon the latest developmentsin the field. The poor elastic recovery of TPEs at elevated temperature is a keydeficiency that has prevented these materials from completely replacing theirthermoset counterparts.
Thermoplastic elastomers owe their existence as products of commerceto the fabrication economics and environmental advantage they offer overthermoset rubber. TPEs, of course, are designed to flow under the action ofheat; hence their upper service temperature is limited in comparison tothermoset rubber. Thus a major hurdle to overcome in the replacement ofthermoset rubber with TPEs is the improvement in elastic recovery, partic-ularly at elevated temperature, especially compression set, because in manyapplications elastomers are subjected to compression. The scope of thischapter includes those TPEs that in our opinion come reasonably close inproperties to thermoset elastomers, as listed in Table 1. Not included, forexample, are plastomers that are ethylene/a-olefin copolymers generallyproduced using metallocene catalysts (5).* These materials can be rubberlikeonly at room temperature. They are thermoplastic owing to the thermorever-sible cross-links provided by crystallization of the ethylene sequences in thepolymer but are deficient in elastomeric character above room temperature orwhen under excessive strain. Thermoplastic elastomers based onmelt-blendedpolyolefins, ethylene/vinyl acetate copolymers, and ethylene/styrene co-polymers are also omitted from the list (6,7). Although thermoplastic olefins(TPOs) represent a commercially important class of materials, they areincluded primarily as comparative points to their more elastomerically per-forming counterparts, TPVs.
Plasticized poly(vinyl chloride) (PVC) is used as a flexible plastic andnot an elastomer but is included in Table 1 because it was the first commer-
cially produced thermoplastic elastomer. PVC, produced by free radicalpolymerization, contains crystallizable syndiotactic segments, the crystalli-zation of which is enhanced on mobilization of the polymer chain in thepresence of a plasticizer (8). However, imperfections in the crystalline phaselimit the upper service temperature of PVC.
II. THERMOPLASTIC ELASTOMERS: APPLICATIONSOVERVIEW
Thermoplastic elastomers are found in thousands of applications, rangingfrom commodity TPOs used in automotive bumper and facia applications,through plastomers used as impact modifiers for plastics, and TPVs and SBCsin sealing applications, to TPUs and copolyesters in numerous engineeringapplications. TPEs replace EPDM rubber in many sealing applications, butylrubber where permeation resistance is required, and nitrile rubber for oil andfuel resistance.
World demand for thermoplastic elastomers will grow at over 6% peryear through 2006, according to a recent study (9). The 1.6 million metric tonTPE industry will remain concentrated in the United States, Western Europe,and Japan, although underdeveloped markets such as Asia grow at a fasterrate.
The most important driver for TPE growth through thermoset rubberreplacement is cost savings. This is normally achieved through a combinationof material selection, part redesign, and fabrication economics. Recyclabilityand weight reduction provide additional drivers in some markets. Colora-bility is another important TPE attribute that increases design flexibility.Further, use of TPEs allows introduction of designs, processes, and value-added features not possible at any cost with thermoset rubber.
Almost all commercial TPEs have one feature in common: they aremicrophase separated systems in which one phase is hard at room tempera-ture while another phase is soft and elastomeric. The harder phase gives TPEtheir strength and, when softened, their processability. The soft phase givesTPEs their elasticity. Each phase has its own glass transition temperature, Tg,or crystal melting point, Tm, and these in turn determine the temperatures atwhich the TPEs exhibit their transition properties. Thus, the TPE servicetemperature on the lower end is bounded by the Tg of the elastomeric phase,whereas the upper service temperature depends on the Tm of the hard phase.Note that the practical service range also depends on the softening point,stress applied, and article design (10).
The ability of TPEs to repeatedly become fluid on heating and solidifyon cooling givesmanufacturers the ability to produce rubberlike articles using
the fast processing equipment designed for the plastics industry. Scrap canusually be reground and recycled. Output of parts is generally increased andlabor requirements reduced compared to parts manufactured from thermosetrubber. Thermoplastic elastomers can be fabricated by conventional thermo-plastic methods including injection molding, blow molding, and extrusion.
Injection molding processes range from single- to multiple-cavity,including up to 48 or more cavities per mold, hot runner mold technologyfor runnerless part production, insert molding with other materials, andcoinjection molding of two materials sequentially or simultaneously. Toolssuch as MoldflowR (11) allow fast development of tooling and processconditions for many TPEs. Another significant advantage is that injectionmolding of TPEs allows dimensional tolerances not achievable in thermosetrubber. This allows snap fits and ‘‘living hinges’’ to be designed into the parts.Flexible, nonblooming, flashless parts are easily produced on largely auto-mated molding equipment. A compatible thermoplastic can give excellentbond strengthwith two-shot injectionmolding. For noncompatiblematerials,a physical lock or interference fit is used over a rigid substrate ofmetal, plastic,or even glass (12).
Blow molding is practiced by injection blow molding, extrusion blowmolding, or press blow molding processes. Complex designs can be easilymanufactured by three-dimensional sequential blow molding with multiplematerials. Fabrication process equipment is available today that can blowmold three-dimensional parts from combinations of thermoplastic and ther-moplastic elastomer materials in up to seven layers by precise material de-livery, robotic parison manipulation, and perfectly timed mold positioning,all computer-controlled in a largely automated process (13).
Extrusion of thermoplastic elastomers includes single-extrusion, co-extrusion, and triple-extrusion processes. Multiprofile dies for extrusionsfrom a single line provide important improvements in efficiency for simpleextrusions. Hard–soft combinations with other polymers, including polyole-fins, polystyrene, and other TPEs, are commonly practiced. Recent develop-ments include coextrusion of thermoset EPDM with TPVs (14,15). Specialextrusion processes have been developed to produce foamed profiles usingwater as the blowing agent (16,17) and create low-friction surfaces with acoextruded slipcoat, offering low-cost environmentally friendly alternativesfor specific applications. Robotic extrusion of TPVs, through a systemcomposed of a moving die, flexible heated hose, and 3D robot, has been usedto apply seals directly to automotive parts (18,19). Secondary processes suchas heat welding, thermoforming, coating, printing, and painting add signifi-cant value at moderate cost in many applications.
Thermoplastic elastomers can offer the design engineer greater designflexibility as well as part size and weight reduction. In the case of thermoset
rubber replacement, the part is usually redesigned to leverage the physicalproperties and processing characteristics of the TPE. The use of TPEsfrequently allows designers to reduce the amount of material per part and,combined with the lower specific gravity of TPEs in comparison with thermo-sets, significantly reduce the overall part weight compared with thermosetrubber (20). An important advantage in redesign is the opportunity for partsconsolidation through combinations of thermoplastic elastomer and otherthermoplastic components.
Thermoplastic elastomer grades have been developed that bond to awide range of engineering thermoplastics, including polypropylene, polyeth-ylene, polystyrene, polyamides, polyesters, acrylonitrile/butadiene/styrene(ABS) rubber modified plastic, cured EPDM rubber, polycarbonates, andcopolyesters. The bond is typically formed through an autoadhesion (diffu-sion) mechanism during thermoplastic processing (21,22). In many cases,bond strengths at levels comparable to material strength can be achieved.
A. Thermoplastic Elastomers in Automotive Applications
The automotive industry has always been a major end-use market for TPEsand accounts for about 60% of the total demand in North America. Tiresaccount for most of the thermoset elastomeric content in a vehicle. The rest isspread over 600 or more elastomer applications from simple grommets tocomplex constant-velocity joint boots and radial lip seals. Automotiveelastomeric parts serve in a wide range of operating environments. They alsoprovide numerous functions such as air, vacuum, and fluid seals; mechanicalshock absorption; flexible couplings; and soft-touch interior components. Aswith any elastomer, TPEs have their limitations. They do not have thecombination of abrasion resistance, flexural strength, deformation resistance,and high-temperature use that thermoset elastomers display; therefore, thesematerials have found no significant use in pneumatic tires.
Key automotive trends have provided a demand for increasing use ofTPEs. The most important is the drive for cost reduction in every possiblecomponent of the vehicle. Even though TPEs are more expensive as a rawmaterial than thermoset elastomers, the cost of the TPE finished part isusually significantly lower than that of a functionally comparable thermosetrubber part through redesign including lighter weight, shorter cycle time,lower energy usage, lower scrap, and recyclability.
Another significant automotive trend is the increased level of govern-ment regulations, which has forced the world’s automotive manufacturers toput major emphasis on improving safety and increasing fuel efficiency,recyclability, and the use of environmentally friendly materials. As Germanyled the world in reduction of nitrosamine-containing cure package compo-
nents for thermoset rubber, the European Union leads with respect tolegislation requiring higher recyclable content and lower overall vehicleemissions (23). Recyclability has provided a consistent driver in the Japanesemarket. Thermoplastics and thermoplastic elastomers are key to reaching thetarget (24). Vehicle manufacturers have taken a lead as well, including targetsfor increase in recyclable content and elimination of PVC use in certain autointerior skin applications. The relatively low price of PVC compounds,however, makes replacement by olefinic systems difficult from a cost view-point (25).
In addition, the automotive industry is trying to respond effectively toan increased level of technical performance requirements. Higher perfor-mance engines, operating at higher temperatures with lowered emissions,coupled with improved aerodynamics due to decreased frontal and grille area,contribute to increasing under-the-hood temperatures. Longer lived automo-biles also require elastomers with improved ultraviolet resistance. Soft-touch,color-matched interior parts, featuring low odor and low fogging, add toesthetics and consumer-recognized value.
Engine compartment timing belt covers with a flexible segment ofrubber and a rigid segment of polypropylene have successfully employedTPE. Fuel line covers from specially formulated flame-retardant grades, rack-and-pinion boots taking advantage of the outstanding flex fatigue resistanceof TPVs, and clean air ducts featuring innovative convolute designs incombination with polypropylene are just a few examples of automotiveapplications that leverage the unique properties of TPVs. Thermoplasticelastomers, especially thermoplastic vulcanizates, are moving quickly intoautomotive weatherseal applications; this market provides significant growthpotential for TPEs in the future. TPEs are injection molded for glassencapsulation and cutline seals. They are extruded for belt line and glassrun channel seals. Extruded seals can be coated with specially formulated lowfriction TPEs and joined at the corners with specialty molding TPVs toreplace flocked thermoset EPDM seals with 100% recyclable parts.
B. Thermoplastic Elastomers in Industrial Applications
Thermoplastic vulcanizates are found in hundreds of industrial applications.In most cases the drivers for TPE use are the same as in other industries, i.e.,thermoset elastomer performance with the advantages of thermoplasticeconomics. The building and construction industry takes advantage of TPEperformance to provide critical sealing in places such as architectural glazingseals, bridge deck seals, pipe seals, and roofing. Industrial hose applicationsform a growing segment of TPV applications, including fire hose, washdownhoses, and specialty grades for handling potable water and food. Excellent
TPV resistance to detergents, acids, and bases, combined with superior flexlife and weatherability compared to thermoset rubber, drive application inthousands of small sealing parts such as gaskets and bushings in appliancesand mechanical devices worldwide. Specialty TPEs featuring low flameretardancy, good abrasion resistance, dielectric strength, and wet electricalperformance are used in electrical applications, especially wire and cablecoverings, insulators, and flexible connectors (26). Conductive thermoplasticelastomers incorporating carbon or metal powders are used for staticdissipative and conductive properties or in electromagnetic interference/radiofrequency interference (EMI/RFI) shielding (27).
Multilayer coated sheets are used in roofing, and their use is expandingto innovative applications such as pillow tank liners.
C. Thermoplastic Elastomers in Consumer Applications
Thermoplastic vulcanizates are found in a variety of consumer products, mostrecognizably those incorporating grips for soft but secure handling of powertools, housewares, and toothbrushes. Good sealing properties and goodchemical resistance make them well suited for kitchen appliances (28).Because many TPEs have consistent frictional characteristics over a rangeof temperatures and in wet and dry conditions, they are well suited for use inthis growing market. The ability to adhere to a variety of substrates by two-shot or overmolding allows processing ease with excellent adhesion. Trans-parent and translucent products are readily available.
Many ballpoint pens now feature a soft grip made from a TPE.Cosmetic containers, food containers, and water bottles incorporate TPEsfor soft-grip feel, color, and design innovation. The demand for thermoplasticrubber soft grips is also growing in sports applications, such as tennis racketor golf club grips. Other sports and leisure applications include toys, skiequipment, and sports balls (e.g., soccer ball inner bladder) made from butylrubber–based TPVs. Consumer products emphasize good esthetic design aswell as functionality, and the ability of TPEs to be decorated is a realadvantage. Techniques such as permanent laser marking and the applicationof hot stamping foils, heat transfer labels, or screen or tampo printing havebeen used for marking various products, including multicolored flexiblelabels. Logos can be integrally designed into products by using overmoldingof hard–soft combinations. Effects linked to other materials such as mineralscan be obtained through the use of innovative pigments; marble and graniteare the most commonly imitated materials (29). Newer application areas forTPEs in consumer products include personal electronics and a growing rangeof household and garden tools.
The segmented block copolymer TPEs included in Table 1–3 contain sequen-ces of ‘‘hard’’ and ‘‘soft’’ blocks within the same polymer chain. Solubilitydifferences between the polymer segments and association and/or crystalliza-tion of the hard blocks produce phase separation in the molten elastomer as itcools. The hard blocks form the thermoreversible cross-links and reinforce-ment (increasing stiffness) of the elastomeric soft phase. The rate of crystal-lization or association of the hard blocks will impact product fabrication time.Polymer microstructure and morphology is depicted in Figure 1. These TPEsare produced by condensation or addition step growth polymerization andhave low molecular weight segments. Although this is desirable, segment mo-lecular weight andmolecular weight distribution cannot be readily controlled.In a 40 Shore D copolyester (COPE) elastomer based upon poly(butyleneterephthalate) (PBT) hard blocks and poly(tetramethylene oxide/terephthal-ate) (PTMO-T) soft blocks, the hard sequence length varies from 1 to 10 (30).PBT molecular weight of sequence length 10 is 2200, whereas high molecularweight PBT that is commercially available could easily have anMn of 50,000!Thus, a sufficient number of hard blocks have to associate to produce a highenough melting crystal phase to provide a reasonably high elastomer upperservice temperature. This necessitates increasing the hard-phase content of theTPE, which results in a hard elastomer (‘‘filler’’ effect). Note that for a givenhard-phase content, the lower the number of hard domains (more hardsegments per domain), the greater the entropic penalty imposed on theelastomeric phase and the less favored the phase-separated morphology.
Increased hard phase content also causes more hard segments to berejected into the amorphous elastomeric phase, thus raising the rubber glasstransition temperature (Tg) and therefore also the TPE lower service temper-ature. In the case of an increase in the number of hard domains, the soft-phaseTg is also elevated owing to the increased ‘‘cross-link density.’’ Theseconsiderations allow the commercial viability of only hard COPEs. This isa major deficiency in this class of TPEs as the softest product available has ahardness of 35 Shore D. Also based on the above discussion, the more or lesscontinuous hard phase in commercially available COPEs where fibrillarcrystalline lamellae (due to short hard segments) are connected at the growthfaces by short tie molecules can readily be rationalized. The amorphous phaseis also continuous (31).
It is difficult to produce useful soft elastomeric products from segmentedblock copolymers except in the case of thermoplastic polyurethanes (TPUs).The strong association of hard blocks even at low hard block content allowsthe preparation of soft elastomeric TPUs. TPUs with hardness as low as 70Shore A are available commercially.
Thermoplastic polyurethane (TPU) was the first thermoplastic product thatcould truly be considered an elastomer (32). The bulk of commercially avail-able TPUs are produced from hard segments based on 4,4V-diphenylmethanediisocyanate (MDI) and 1,4-butanediol (BDO, a ‘‘chain extender’’), witheither poly(tetramethylene oxide) (PTMO) glycol, or poly(1,4-tetramethyleneadipate) (PTMA) glycol or poly (q-caprolactone) (PCL) glycol as the softelastomeric segment (32). TPUs can be produced by a ‘‘one-pot’’method or in
Figure 1 Polymer microstructure and morphology of segmented block copolymers(TPU, COPE, COPA). A, crystalline domain; B, junction area of crystalline lamella;C, polymer hard segment that has not crystallized; D, polymer soft segment.
a two-stage process. In the former, the diisocyanate, chain-extender diol, andsoft segment diol are mixed and heated to yield the final product, whereas inthe latter the soft-segment diol is first ‘‘end-capped’’ by using an excess ofdiisocyanate and the chain-extending short-chain diol is subsequently addedto form the hard segments and to attach them to the soft segments in analternating manner to yield a TPU of high molecular weight by addition step-growth polymerization. A representation of a TPU molecule is presented inFigure 2. A TPU’s Mw can be as high as about 200,000, with Mn about100,000, although the individual hard and soft segments are of much lowermolecular weight. For example, poly(tetramethylene oxide) glycol ofMn 1000or 2000 is used commercially for TPU production, thereby fixing the softblock length. The longer the soft segment, the lower its hydroxyl end groupconcentration, which would allow preferential step growth of the hardsegments by reaction of the short-chain diol with the diisocyanate. Hence,the longer the soft segment, the longer the hard segment. Because the numberof soft segments will equal the number of hard segments, for a large number ofalternating segments,
where soft segments and hard segments are abbreviated SS and HS, respec-tively. For given soft segment molecular weight, the number-average molec-ular weight of the hard segment is directly proportional to the hard segmentcontent and inversely proportional to the soft segment content (33).Mwss canbe obtained by measurements on the polyol, but obtaining the hard-segmentweight-average molecular weight is difficult. Bonart developed a theoreticalmethod to calculate Mwhs (34). The average number of hard segments for aTPU (MDI/BDO hard segments; polyoxypropylene end-capped with poly-oxyethylene soft segments) with a 50 wt% hard phase has been calculated tobe six (35). Peebles mathematically modeled the soft and hard segment lengthdistribution in TPUs (36,37).
The infrared studies of Cooper demonstrated that the urethane NUH ishydrogen-bonded to the oxygen atoms of the urethanemoiety as well as to theoxygen atoms of the polyether or polyester soft segments (38). This hydrogenbonding and soft segment polarity can retard and lower the ultimate degree ofphase separation in TPUs. Poor phase separation is reflected in the increase inTg of the mostly amorphous soft phase due to the presence of dissolved hardsegments. The hard microphase is formed by association of the relativelyshort hard segments and by their crystallization into fibrillar microcrystals.The poorer phase separation in polyester TPUs compared with polyetherTPUs is presumably due to the greater polarity of and stronger hydrogenbonding (with the NUH of the hard segments) in the soft phase of the formercompared with the latter (39). A 1:2:1 (molar polyester:MDI:BDO) TPU(polyester polyolMn=1000) exhibited a single phase, but the correspondingpolyether-based TPU system was phase-separated (40). The degree of phasemixing is also dependent upon soft segment content. For a polyether-basedTPU, complete phase mixing was observed at 80 wt% soft segment content(41,42). Phase mixing is also dependent upon segment molecular weight, asdemonstrated in the case of TPUs containing low molecular weight poly-caprolactone soft segments (43).
Phase separation in TPUs is driven by the solubility parameter differ-ence between the polymer segments and by association and/or crystallizationof the hard segments and is limited by the geometry of the molecule and thehydrogen bonding and polarity effects discussed. In addition, the kinetics ofTPU phase separation will also be influenced by the mobility (Tg) of thepolymer segments.
A. TPU Morphology and Microstructure
The mechanical behavior (Young’s modulus, elastic recovery, elongation,flexural modulus, heat sag, thermomechanical penetration probe behavior) ofTPUs suggests a transition from discrete to continuous hard microdomain
morphology at hard segment content above about 45 wt% (33,41–46). Thesmall-angle X-ray studies of Abouzahr and Wilkes (42) and Cooper andcoworkers (43) and the small-angle X-ray and neutron scattering analysis ofLeung and Koberstein (41) suggested an interlocking hard domain morphol-ogy at high hard segment content.Depending upon processing conditions andhard phase type and content, crystalline TPU systems may exhibit a fringedmicellar texture of thickness equal to the hard segment length or clear-cutconnectivity of the crystalline hard phase. The hard domain diameter in aTPU produced from a 1:6:5 polycaprolactone (Mn=2000)/MDI/BDOmoleratio was estimated to be 400 A by transmission electron microscopy (TEM)(46) (hence ‘‘hard microdomain’’), although for the typical TPU materialsmentioned in this review this number is expected to be about 100 A.
Using small-angle X-ray scattering (SAXS), Leung and Koberstein (41)studied the hard segment microdomain thickness (which corresponds to thelength of the hard segments) in TPUs in which the hard segment contentvaried from 30 to 80 wt%. The SAXS measurement provided an overallcharacterization of the microdomain morphology averaged over crystallineand noncrystalline structures. The hard microdomain thickness varied from2 nm (corresponding to a hard segment length containing twoMDI residues)to 5.4 nm (hard segment length with four MDI residues) for the 60 wt% hardsegment content TPU, after which the thickness did not increase further withincreased TPU hard segment content. Because the hard segment lengthincreases with increased TPU hard segment content, chain folding via theflexible BDO segments to accommodate longer hard sequences within thecrystal is thought to occur. Other possible explanations for this phenomenonhave been discounted. The extended chain crystal structure, irrespective ofTPU hard segment length, that has been demonstrated to occur by wide-angleX-ray diffraction (WAXD) may well be characteristic of the TPU samplesstudied that were treated (annealed, etc.) to maximize crystallinity so as to beamenable to analysis by the WAXD method (41).
Spherulitic structure for high hard segment content (>40 wt%) TPUshave been observed in samples crystallized in the laboratory (33,46,47). In onecase, because of the large spherulite diameter (several micrometers) and theabsence of a hard phase Tg, the spherulites may have contained occluded softphase (33). Hard phaseTg is rarely discernible even in high hard phase contentTPUs. A hard phase Tg was observed in a melt-quenched TPU with 80 wt%hard segment content (33). Owing to the tendency of the relatively short TPUhard segments to associate or crystallize or to be miscible in the TPU softphase, amorphous hard segments may exist only as tie molecules connectingmicrofibrillar crystalline segments. Low TPU amorphous hard phase contentwould preclude Tg detection. Moreover, hard phase Tg observation would beobscured by other transitions (discussed later). Spherulitic soft segment
structure in a high PTMO soft segment content TPU has been observed (33).Generally, TPU parts that are fabricated by commercial processing equip-ment exhibit crystallinity but no spherulitic structure (48).
B. Thermal Characteristics of TPUs
Although the structure of TPUs changes constantly during differentialscanning calorimetry (DSC), DSC coupled with SAXS has proven to be apowerful tool in uncovering TPU microstructure and thermal behavior, as inthe masterful research work of Koberstein and coworkers (35,41,49,50), whostudied polyether TPUswithMDI/BDOhard segments.Molten TPUs from ahomogeneous melt state were rapidly quenched to and held at variousannealing temperatures for specific time periods. Generally, three distinctendotherms were observed by DSC of the annealed samples. The firstendotherm (TI) is dependent upon the annealing temperature, annealingtime, and TPU hard segment content. This endotherm is observed at 20–40jCabove the annealing temperature,whichwas varied from30jC to 170jC,depending upon TPU hard segment content. Higher hard segment contentTPUs gave higher TI values. The exact origin of TI is still unknown, but it islinked to a short-range order dissociation endotherm in the hard microphaseand not in the interphase, because this transition is also observed in pure hardsegment materials as suggested by Cooper and coworkers (51,52). For a softTPUwith a discrete hard phase and a total hard phase content of 30 wt%, theTg of the soft phase kept increasing with increased annealing temperature upto 170jC.Annealing above 170jCdid not change the soft phaseTg, indicatingthat the microdomain structure is completely disordered above this temper-ature (35). The Tg increase of the soft phase was related to increasedsolubilization of hard segments into the soft phase. Increasing annealingtemperature caused the solubilization of hard segments of high molecularweight into the soft phase that already contained lowermolecular weight hardsegments. It has also been suggested that ‘‘cross-linking’’ by soft segment–hard segment hydrogen bonding is another factor that contributes to in-creased soft phase Tg in addition to the physical presence of TPU hardsegments in the soft phase (53). By studying the change in TPU heat capacityat its glass transition temperature, it was concluded that below an annealingtemperature of 80jC hard segment solubilization into the soft phase occursand above 80jC, which is near the hard segment Tg, soft segments that aretrapped in the hard microphase also enter the bulk soft phase in addition tofurther hard segment dissolution into the soft phase.
The TII endotherm is also dependent upon annealing temperature, andfor the soft TPU under discussion the TII maximum is 175jC. This transitionwas identified by Koberstein as the microphase separation transition (MST),
where the partially ordered ‘‘noncrystalline’’ segments in the hard micro-domain are mixed into the soft TPU phase. The TPU with 30 wt% hardsegment content did not exhibit a microcrystalline melting TIII endotherm,which is observed for higher hard segment content TPUs. The identificationof TII as the MST was further confirmed by simultaneous DSC/SAXSmeasurements in a TPU with 50 wt% hard segment content (49). The TPUinterdomain spacing increased dramatically beginning at TII. This TPUexhibited a higher TIII endotherm corresponding to the melting of a micro-crystalline hard phase within the ‘‘noncrystalline’’ ordered hard domain.
For the TPU with 50 wt% hard segment content, the TI endothermmerged with the TII endotherm when annealing took place at 155jC.Annealing above 155jC raised the TII endotherm and decreased its intensitywhereas the intensity of the TIII microcrystalline peak melting endothermincreased. TIII was the only DSC peak endotherm observed at 210jC whenannealing was conducted at 175jC. At annealing temperatures of 175–190jC,the TIII endotherm diminished in magnitude and the TII endotherm reap-peared. These findings are consistent with an expected decrease in crystallinityat low undercoolings where crystallization is controlled by nucleation. Abovethe MST, TPU crystallization occurs from a homogeneous mixed melt phase(‘‘solution’’ crystallization). Crystallization occurs within the hard micro-domains (‘‘bulk’’ crystallization) below the MST. For harder TPUs (70 wt%hard segment content), melting endotherms corresponding to different crystalstructures have been observed, depending upon annealing conditions.
The thermogravimetric analysis (TGA) trace of the TPUs of the Hu andKoberstein study (50) demonstrates initial weight loss around 300jC, which iswell above the annealing temperatures used to probe the TPUmicrostructure.A small change in annealing temperature (from 190jC to 195jC) exhibited adramatic increase in TPU Mn and Mw values [gel permeation chromatogra-phy (GPC) measurements]. The increased MW is presumably the result of‘‘trans urethanation’’ reactions that result from cleavage of the urethane bondin a polymer segment back to the isocyanate and alcohol, and subsequentallophanate formation by addition of the newly formed isocyanate to theurethane NUH bond of another polymer chain, thus creating a branchedstructure. Crystallization of the branched TPUmolecules appears to be hind-ered in comparison with their linear counterparts. Reduction in the heat offusion is observed for TPU samples where molecular weight was increased byannealing at high temperature, due to ‘‘trans-urethanation’’ reactions. Itshould also be reiterated here that the sequence length of the hard segmentsthat are incorporated into the soft phase increases with increased annealingtemperature. For more on trans-urethanation reactions and TPU thermaldegradation mechanisms, the reader is referred to the work of Macosko andcoworkers (54).
According toKoberstein, all three TPU endothermsTI,TII, andTIII areaccompanied by the mixing of hard and soft microphases. The Kobersteinschematic model for the morphological changes that occur during the DSCscans of TPUs is presented in Figure 3. It should be noted that Koberstein’swork is grounded on the pioneering TPU research work of Wilkes andCooper and coworkers, who had previously recognized the time- andtemperature-dependent morphological and mechanical properties of TPUs(51,55–61). The increased mutual solubility of TPU hard and soft phases withincreasing temperature was recognized, as was the influence of hydrogenbonding and soft phase Tg on phase mixing and demixing over a broad temp-erature range. Both phase mixing and demixing have been observed on TPUmechanical deformation, depending upon sample thermal history, includingchanges in phase continuity (59). TPU morphology is complex, and a smallchange in the polymer segment type can result in diverse melting behavior.For example, TPUs produced fromMDI/BDOhard segments and poly(hexa-methylene oxide) soft segments exhibited five melting endotherms that wereattributed to hard segment sequences containing one to five MDI-derivedunits (62). There is continued interest in elucidating the origin of multiplemelting endotherms in TPUs (63).
It is now readily understood how TPU morphology is dependent uponprocessing conditions and what thermally induced phase transitions canoccur that would be detrimental to product elastic recovery at elevatedtemperature.
Based upon the information presented so far, it would appear that TPUsthat are designed for improved phase separation (decreased hard and softphase compatibility) should provide improved elastic recovery. However,TPU mechanical properties are adversely affected when the desired micro-structure is difficult to achieve due to incompatibility of the TPU buildingblocks under the polymerization conditions, including incompatibility of thereactants with the polymer produced. This is the case for TPUs (for improvedhydrolysis resistance) produced with polybutadiene diol or hydrogenatedpolybutadiene diol (for improved heat and hydrolysis resistance) soft seg-ments and MDI-based hard segments (64–68). Molecular heterogeneity inchemical composition and average hard segment length is expected to be thekey factor contributing to the poor mechanical properties of these hydrocar-bon soft segment TPUs compared with conventional TPUs, based on, forexample,MDI/BDO/PTMO (69–71). Hydrocarbon diols are being promotedfor nonelastomeric polyurethane applications, as in the preparation ofcastable polyurethanes for moisture-resistant adhesives, coatings, and elec-trical potting compounds (72).
Thermoplastic polyurethanes produced with 2,6-toluenediisocyanate(2,6-TDI) hard segments with BDO as chain extender and PTMO as the soft
Figure 3 Schematic model for the morphological changes that occur during DC
scans of polyurethane elastomer (a) below the microphase mixing transition temp-erature, (b) between the microphase mixing temperature and the melting temperature,and (c) above the melting temperature. The microcrystalline hard-segment domainsare indicated. (From Ref. 49.)
phase undergo cleaner phase separation than the corresponding 2,4-TDIbased TPUs (53). The use of 2,6-TDI as the hard phase isocyanate mayprovide TPUs with excellent elastic recovery, but difficulty in 2,6-TDI/2,4-TDI isomer separation makes this approach commercially unfeasible. More-over, the volatility of TDI over MDI makes the latter isocyanate preferablebecause of toxicity considerations. However, TDI, the first isocyanatedeveloped for the thermoset polyurethane industry, is still used in NorthAmerica in the manufacture of thermoset polyurethane foam (73–75). TPUsproduced with aromatic diol chain extenders such as hydroquinone bis(2-hydroxyethyl) ether in, for example, the conventional MDI/PTMO systemare emerging as elastomers with improved elastic recovery (76).
Aliphatic and aromatic diamines can be used as chain extenders to formTPU ureas with high melting point hard segments, but these materials meltwith some decomposition andwell above the processing temperature of TPUs(32) and hence are not commercially feasible as thermoplastic elastomers withimproved elastic recovery.
However, owing to improved elastic recovery after high strain and ahigher use temperature due to the urea hard segments, solution-processedaromatic polyurethaneureas are preferable to conventional melt-processedaromatic polyurethanes in fiber applications (clothing, upholstery, andcarpet). Spandex is the generic trade name given by the Federal TradeCommission to synthetic elastomeric fibers that contain at least 85% seg-mented polyurethane. In comparison with natural rubber threads, Spandexfibers are readily dyeable, lightweight materials with excellent abrasionresistance, tensile strength, and tear strength. They have better resistance tooxidation, sunlight, and dry cleaning fluids than natural rubber threads andare also tolerant to bleach containing a low chlorine level. Although curednatural rubber fibers have the advantage of low hysteresis and stretchcrystallinity, they are being replaced by Spandex, which can also be curedduring the fiber-forming process (77).
C. Aliphatic TPUs
Aliphatic TPUs are used in light-stable (nonyellowing) applications and canhave mechanical properties comparable to those of aromatic TPUs (78).These materials are synthesized from hydrogenated MDI diisocyanate/BDOor hexamethylenediamine diisocyanate/BDO hard segments and polyestersoft segments. (Polyether soft phase would reduce TPU UV resistance.)Conventional MDI-based aromatic TPUs yellow on exposure to UV lightowing to the formation of quinone imides. The quinone imides are UVabsorbers that dissipate UV energy as heat and hence retard further TPUdegradation. On UV exposure, the aliphatic TPUs undergo a greater reduc-
tion in mechanical properties than their aromatic counterparts but withoutcolor change or loss of transparency. Hence, UV-stabilized aliphatic TPUsare used in outdoor applications where the abrasion resistance of TPUs isnecessary. For example, some outdoor signs enclosed in transparent acrylicare laminated with aliphatic TPUs. Aircraft canopies are fabricated withhigh-impact-resistant layered structures produced from polycarbonate and a‘‘flexibilizing’’ aliphatic TPU ‘‘glue.’’
As illustrated by the data in Table 3, the compression set of TPUs ismuch poorer than that of thermoset rubber. Under compression at elevatedtemperature, irreversible deformation in TPUs occurs by continued phaseseparation and/or reorganization of the hard and soft segments over thatestablished after part manufacture.
Hydrogen bonding in the hard phase and in the interphase (the regionwhere the polymer composition changes from 100% hard segment to 100%soft segment) between the hard and soft domains provides a readymechanismfor chain slip because hydrogen bonds can reorganize readily by the partialformation of ‘‘new’’ hydrogen bonds as the ‘‘old’’ hydrogen bonds arepartially broken. Increasing the amount of the hard phase (to provide moresecure thermoreversible cross-links at the TPE upper service temperature)increases compression set because the now higher modulus material issubjected to much higher stress under compression compared to thecorresponding softer material (under constant deflection). Increased hardphase volume fraction in TPUs also restricts polymermotion in the soft phase(increased elastomer cross-link density), and there is an increased presence ofhard segments in the soft phase. These factors cause an increase in the softphase Tg that raises the product’s lower use temperature. The hard TPUproduct, of course, would have an advantage in constant load applications.
Thermoplastic polyurethanes may also contain thermoreversible allo-phanate branch points resulting from the reaction of the urethane NUHbondwith excess diisocyanate. It is not feasible to design allophanate bonds intoa TPU, but these fortuitously present cross-links may contribute to improvedTPU elastic recovery. Nevertheless, elastic recovery in the various types ofTPUs does not approach that of thermoset rubber. In some cases the elasticrecovery of a soft product can be worse than that of a harder product becauseof product design. For example, it may be necessary to produce a soft TPUwith a low rate of crystallization to achieve desirable processing character-istics in film applications. This may be accomplished by the use of a lowmolecular weight soft segment in which the TPU crystallization rate islowered owing to increased phase mixing. Continued phase separation inthe finished product is one factor that would raise set.
Amorphous materials exhibit a gradual decrease in viscosity withincreasing temperature beyond Tg, compared with crystalline materials, in
which viscosity drops sharply on melting due to the Tm being much greaterthan theTg. In crystalline hard phase TPUs the viscosity drop on crystal phasemelting may not be as precipitous as expected because of association amongthe hard phase molecules that are still present just after melting because ofincompatibility with the soft phase. Even so, this viscosity drop in a crystallinehard phase TPUmay cause it to lack desirable processing characteristics, andTPUs with a high amorphous hard segment content may be designed for animproved processing window and for transparency. The excellent impactproperties, processability, and transparency of Dow’s IsoplastTM are creditedto the amorphous hard segment that makes up most of this TPU engineeringplastic. In the finished product, elastic recovery is controlled by both rawmaterial properties and part design.
V. ELASTOMERIC COPOLYESTERS AND COPOLYAMIDES
Elastomeric copolyesters (COPEs) (31) and elastomeric copolyamides(COPAs) (79) are similar in structure to TPUs and suffer similar drawbacksin rubber performance. The hydrogen bonding present in TPUs andCOPAs isabsent from COPEs. Commercially available COPEs are based upon crys-talline polybutyleneterephthalate (PBT) hard segments and poly(tetramethy-lene oxide) (PTMO) soft segments. PBT monofilaments exhibit only a 1%permanent set after 11% extension at room temperature, owing to a reversiblea- to h-crystal transition (80,81). This reversible crystal transition, whichwould be beneficial in the elastic recovery of COPEs, has been observed inPBT/PTMO COPEs with a high enough level of the PBT hard phase that theamorphous phase is hard enough (due to the presence of PBT hard segmentsin the amorphous phase) to bear the level of tensile stress necessary to causethe reversible deformation behavior in the hard phase (82). Although it isgenerally thought that segmented block copolymers have a homogeneousamorphous phase consisting of hard and soft blocks, experimental evidenceindicates that a biphasic amorphous phase consisting of a PTMO phase and amixed PBT/PTMO phase can exist in certain COPEs (83,84). The lack ofhydrogen bonding in COPEs and the reversible crystal transformationpossible in the PBT hard phase are responsible for the modest improvementin elastic recovery of these materials over TPUs and COPAs. However, atelevated temperature, the motion of the soft segments cannot be adequatelyrestrained by the crystalline polymer chains, thus causing reorganization inthe hard phase that leads to irreversible deformation. COPEs cannot matchthe elastic recovery of thermoset rubber (Table 3).
In addition to the disadvantage of poor elastic recovery at elevatedtemperature that is characteristic of most TPEs, the segmented block copoly-
mers suffer the additional disadvantage of the lack of commercially availablesoft products due to inadequate physical properties as already discussed. Inthe case of TPUs, the association of the hard segments is strong enough toconfer excellent physical properties to soft products (70 Shore A), butdifficulty in pelletization of the soft product during manufacture and pelletagglomeration on storage have to be overcome.
Addition of plasticizer to hard segment block copolymers is not a viableoption for the production of soft products, because the plasticizer wouldlower the melting point of the polar hard phase in addition to softening thepolar elastomeric phase, which, in any case, cannot hold a high level of addedplasticizer. Moreover, continued phase separation after processing can causethe exudation of plasticizer from themolded product. Commercially availablesegmented block copolymer TPEs are plasticizer-free.
Elastic recovery is an important property for elastomer performance.Because of the price and performance requirements in diverse applications,the hydrocarbon oil-resistant segmented block copolymers discussed aresuccessful products of commerce.
The most important end use of the polyurethane-elastomer, polyamide-elastomer, and polyester-elastomer block copolymers has been in thermosetrubber replacement. Their crystalline hard segments make them insoluble inmost liquids. Products feature exceptional toughness and resilience, creep andflex fatigue resistance, impact resistance, and low-temperature flexibility. Allthree types are generally used uncompounded, and the final parts can bemetallized or painted. Thus, they are often used as replacements for oil-resistant rubbers such as neoprene because they have better tensile and tearstrength at temperatures up to about 100jC.Automotive applications includeflexible couplings, seal rings, gears, timing and drive belts, tire chains, andbrake hose. Special elastomeric paints have been developed that match theappearance of automotive sheetmetal; such parts have been used in car bodies(31,32,79). Flexible membranes, tubing, hose, and wire and cable jackets areincluded in the long list of applications.
VI. STYRENIC BLOCK COPOLYMERS
The advent of hydrogenated styrene/butadiene/styrene (SBS), i.e., styrene/ethylene-1-butene/styrene (SEBS), triblock copolymer compounds repre-sented an advance in the elastic performance of thermoplastic elastomers atelevated temperature. SEBS is almost always compounded; one can achieveprocessable soft compositions (0–30 Shore A) that are not possible in the caseof segmented block copolymers. The key features of SEBS will be describedbefore we discuss SEBS compounds. Phase separation in these triblock
copolymers is more complete and occurs more readily than in the segmentedblock copolymers. This is reflected in the Tg of the rubber phase, which isnearly unaffected by the polymer styrene content. The Tg of the styrene phasedepends upon itsmolecular weight.More phasemixing with the rubber can beexpected with decreasing styrene molecular weight when the material isheated to the Tg of styrene (85). Both the polystyrene end block contentand polystyrene molecular weight in SEBS is designed to be lower than that ofthe rubber midblock. For example, Kraton G1651(SEBS) of Kraton Poly-mers has a plastic block of molecular weight 29,000 (33 wt%) and a rubberblock of molecular weight 116,000 (68 wt%) (86). The rubber block isdesigned to have a 40 wt% butene content to limit crystallinity due to thepolyethylene segments (low crystallinity would increase the rubber’s oil-holding capacity) and lower Tg (low Tg for improved low-temperatureperformance) (87). Simplistically, SEBS has a ‘‘spaghetti and meatball’’morphology, in which the styrenic microdomains (200–300 A) are dispersedin a continuous rubber matrix (88). The polystyrenemicrodomain size reflectsthe entropic penalty that would be imposed on the rubber in the case of largerplastic domains. The higher molecular weight and narrower molecular weightdistribution of SEBS than those of the segmented block copolymers arefactors that favor improved phase separation in the former system in spite ofthe smaller solubility parameter difference between the phases in SEBS versusthe segmented block copolymers (89,90). Molecular architecture also favorsbetter phase separation in SEBS than in the segmented block copolymers. Thepolystyrene phase will flow above its Tg (f95jC), and these microdomainsform the thermoreversible cross-links in the SEBS thermoplastic elastomer.The styrenic cross-links, however, do not contribute much to the ‘‘cross-link’’density of the rubber phase that is dominated by the trapped entanglementswithin it (91). This can readily be inferred by a comparison of the modulus(initial slope of the stress–strain curve and also the plateau modulus) of SEBSwith other styrenic block copolymers such as styrene/butadiene/styrene (SBS)and styrene/isoprene/styrene (SIS). The modulus in these systems is directlyrelated to the molecular weight between entanglements in the rubber phase(88). The modulus of SEBS (lowest molecular weight between entanglementsand highest entanglement density) is greater than that of SBS, which in turnhas a higher modulus than SIS (highest molecular weight between entangle-ments and lowest entanglement density).
Thus the function of the styrenic domains is to prevent disentanglementof the rubber segments when these styrenic block copolymers (SBCs) aresubjected to load. For example,KratonG1651 has a 33.3 wt%PS content anda rubber molecular weight of 116,000 (Mn gMw). Neglecting the interphase,the total PS phase volume in 100 g of SEBS would be 31.71 cm3 (PS density=1.05 g/cm3). Assuming spherical 200 A diameter PS domains, the volume per
domain is 4.19 � 10�18 cm3, which translates to 7.57 � 1018 domains in theSEBS sample. The number of PEB macromolecules is 35.29 � 1019 (68/116,000= 5.89� 10�4 g�mol= 5.86� 10�4� 6.023� 1023 macromolecules).Assuming a molecular weight between entanglements for PEB of 1800, thenumber of entanglements per chain is 64 (116,000/1800). If entanglementsoccur only by the crossing of two different rubber chains, the total number ofentanglements in the rubber is 1129� 1019 (35.29/2� 1019� 64), which resultsin 1490 entanglements in the rubber phase per PS domain. A representation ofSEBS polymer microstructure andmorphology is presented in Figure 4. Notethat in SBS and SIS the rubber block has a high 1,4-copolymerized dienecontent that maximizes phase separation (due to maximized incompatibilitybetween the plastic and rubber phases) for improved elastic properties but isalso detrimental to product processability. On the other hand, SEBS isproduced by the hydrogenation of high- ‘‘vinyl’’ (low 1,4-copolymerizeddiene) SBS for reasons already discussed. Hydrogenation of commerciallyavailable SBS would yield a crystalline plastic instead of an elastomericpolymer midblock.
The foregoing discussion is based upon the ‘‘spaghetti and meatball’’SEBS morphology described earlier. In the case of lower molecular weightSEBS, a higher modulus has been observed compared to those of thecorresponding higher molecular weight counterparts. This has been attribut-ed to the presence of a larger interphase in the former case due to greater phasemixing (92). If the TPE hard block content is high enough to form acontinuous phase, a higher modulus can be expected.
Upon increasing PS content, the discrete plastic phase morphology inSEBS can change to a cocontinuous rubber and plastic phase, and further to adiscrete rubber phase in a plastic matrix. Also, the shape of the plastic phasecan change from spheroidal to cylindrical to plate-like with increasing SEBSPS content. These regular shapes can be achieved only under carefullycontrolled annealing or shearing conditions.
Compared with a corresponding low molecular weight polymer, highmolecular weight SEBS exhibits superior mechanical properties and can be
Figure 4 High rubber content SEBS triblock copolymer microstructure and mor-phology.
used to produce lower cost end products owing to its ability to absorb largeamounts of paraffinic oil. However, high molecular weight SEBS is notprocessable, because this polymer alone does not flow well under polyolefinplastic processing conditions (93). This is due to phase incompatibility thatnecessitates high temperature and high shear (for increased phase-mixingkinetics) conditions to transform biphasic SEBS to a molten single-phasesystem. That is, SEBS has a high order–disorder transition temperature(TODT) that is related to the segmental molecular weight and compositionof this triblock copolymer. For example, the TODT of Kraton G1650 (PSblock 29 wt%, MW= 13,500; PEB block 71 wt%, MW= 66,400), which isconsidered amediummolecular weight product, is estimated to be 350jC (94).Moreover, this transformation would not occur instantaneously at thistemperature; it is expected to be retarded due to the highly entangled natureof the rubber phase. One way of determining the TODT is to experimentallymeasure the temperature at which there is a precipitous drop in polymerelastic modulus when measured as a function of temperature at a fixedfrequency, although this approach may not yield the true TODT (95), becausesome order may still exist in the polymer melt at this temperature. For anexcellent discussion of SEBS TODT the reader is referred to the work of Chunand Han (94), Kim et al. (95), Baetzold and Koberstein (96), and thereferences cited in these publications. In spite of the saturated backbone inhigh molecular weight SEBS, polymer degradation occurs before the TODT isreached, and hence it is difficult to measure this temperature experimentally(94). Lower molecular weight SEBS polymers could be readily processedunder normal polyolefin plastic processing conditions (200–250jC) butcannot provide the necessary price–performance balance to become a productof commerce as an elastomer. A SEBS polymer with a PS end block MW of3400 (31.8 wt%) and a PEB midblock MW of 14,600 (68.2 wt%) exhibits aTODT of 142jC (96).
A. SBCs as Compounded Materials
In elastomer applications, SEBS is never used alone; it is always compoundedto improve product processability and performance and to lower productcost. Polypropylene (PP), paraffinic oil, and fillersmake up the bulk of a SEBSelastomer compound. In elastomer applications, highmolecular weight SEBSis extended with from 200 to over 400 phr of paraffinic oil. In certain oil-gelapplications, the concentration of SEBS is as low as 5 wt% (97). The oilcontributes to compound processability and lowers cost without sacrificingthe elastomer upper service temperature. Paraffinic oil is chosen to selectivelyswell the continuous rubber phase, leaving the discrete polystyrene domainsunplasticized, thereby maintaining the integrity of these virtual cross-links at
the elastomer upper service temperature. The molecular weight of the PS endblocks is high enough to prevent significant plasticization by paraffinic oil andto provide sufficient incompatibility with the rubber phase for balancing TPEelastic properties (better with increased phase incompatibility) with process-ability (better with increased phase compatibility). Because SEBS is producedby the selective hydrogenation in solution of the high vinyl butadiene rubbermidblocks in SBS (98), the narrowmolecular weight distribution of the plasticand rubber blocks in SBS (99) (synthesized by anionic polymerization) ismaintained in SEBS. Thus, a truly uniform rubber network structure swollenin paraffinic oil can be expected for SEBS due to the reorganization possible(at elevated temperature) in the polystyrene domains.
The presence of a uniformly entangled rubber network and the lowinterphase volume (due to polystyrene and rubber phase incompatibility—theinterphase would hold less oil than the rubber phase) expected in highmolecular weight SEBS would explain the large oil-holding capacity of thismaterial. The rubber polymer chains can be viewed as being surrounded by a‘‘tube’’ of oil, where the oil molecules are generally restricted to move withinthe tube but can cross over between tubes. The absorption of oil by SEBS isdriven by the configurational entropy gain by the oil, which overcomes theconformational entropic losses on stretching of the rubber segments. Theremay be some lowering of system internal energy due to the adoption of low-energy conformations by the rubber segments. There also may be a limitedenthalpic attraction between the oil and rubber. The rubber and the oil arenonpolar; therefore, no preferred orientation around the rubber molecule isexpected for the oil in order to maintain the expected enthalpic attraction,thus minimizing the loss in entropy of the oil. There is a slight increase in theTg of SEBS rubber (40 wt% 1-butene) when it is plasticized by paraffinic oil(100).
The viscosity of SEBS drops when it is plasticized by oil, but there is noincreased phasemixing in the ‘‘melt.’’The apparent viscosity is reduced owingto the reduction in frictional forces between the rubber phase (when swollen inoil) and the wall of the capillary rheometer. This friction is not affected muchby shear rate or temperature, so the apparent viscosity varies inversely withshear rate and is almost independent of temperature (92,93). The flow ofSEBS is best described by plug flow resulting from wall slip.
The presence of both polypropylene (PP) and paraffinic oil is requiredfor a dramatic improvement in the processability of a SEBS compound.Molten PP forms the viscous medium that allows ready transport of thebiphasic SEBS during processing. The oil in the SEBS partitions between theSEBS and PP phases (100) (molten PP is miscible with paraffinic oil), thusreducing the viscosity of the molten PP and increasing its volume, whichtranslates into improved SEBS compound melt processability. On cooling,
the molten PP crystallizes, and the oil rejected from the crystalline phasepartitions between the SEBS and amorphous PP phases. On cooling, theSEBS compound ‘‘hardens’’ rapidly due to the crystallization of the PP phase,thereby allowing rapid cycle time in end product manufacture.
B. SBC Morphology
An example of typical SEBS morphology is represented in Figure 4. Themorphology of a SEBS compound is dependent upon the relative proportionsof PP, SEBS and oil and upon processing conditions. Owing to the flowproperties of SEBS already discussed, equilibrium morphology is notachieved in a typical compounding operation, where the residence time inthe extruder is less than 3 min. At a low level of SEBS and oil, particulateSEBS has been found to be dispersed in PP when compounded in a twin screwextruder [75 wt% isotactic homo-PP, MI = 5.5; 13.3 wt% Kraton G1651(33.3 wt% PS, MW = 29,000; 68 wt% PEB, MW = 116,000); 11.7 wt%paraffinic oil]. Polypropylene is the dispersed phase at very lowPP levels, but acocontinuous SEBS and PP phase is present in a composition range fromabout 10 wt% to 55 wt% PP (11.6 wt% PP, 46.5 wt% Kraton G1651, 41.9wt% paraffinic oil). The interdomain movement of the PS segments at thecompound processing temperature causes the high molecular weight SEBS(which is a powder at room temperature) to ‘‘knit’’ together and form acontinuous phase, especially when SEBS forms the bulk of the polymer blend.The paraffinic oil partitions between the SEBS rubber phase (the PS domainswould also absorb a small quantity of oil) and the molten PP. It has beendemonstrated that part of the molten PP, oil, and the PEB rubber phase ofSEBS are miscible, allowing PP to form a continuous phase even at a very lowPP level. The rubber and PP molecules are then entangled, and, on cooling,the trapped entanglements allow good adhesion between the phases and PP isnucleated across the phase boundary so that cocontinuity is maintainedbetween the phases (86,100,101). It is conceivable that the entanglement witha rubbermolecule of an amorphous PP tie chain is anchored if the tiemoleculeis trapped within the same or different PP lamellae as it emerges from therubber phase. Even at high elongations the cocontinuous blends show astress–strain behavior similar to that of rubber, with no sign of the typicalnecking phenomenon normally associated with PP at large deformation. Itseems reasonable to propose that PP is present as thin coiled sheets andligaments that simply uncoil during deformation, so that the PP phase itself isnot subject to much stress and most of the deformation occurs in the SEBS.
From the foregoing discussion it can also be understood how SEBScompounds with a hardness of about 0 Shore A can readily be produced.SEBS can absorb large quantities of paraffinic oil to yield a soft rubber, andthe SEBS compound processability is enhanced by the oil together with a
limited amount of the PP, which forms a cocontinuous hard phase alongsidethe cocontinuous SEBS rubber phase. During processing of the SEBScompound melt it is simply slipping along the processing conduits on a thinfilm of a molten PP solution in oil. High filler and oil loading allows theproduction of low cost SEBS compounds.
Because the oil-holding capacity of SEBS has been discussed, it is worthmentioning the oil-holding characteristics of commercially available SBS inconnection with the wrist rest application, an example of which is a pad thatspans the length of a computer keyboard support. The highest molecularweight linear triblock SBS (Kraton D1101) is medium in molecular weightcompared to the highest molecular weight SEBS that is commerciallyavailable (Kraton G1651) (102). Kraton D can probably hold only 100–150phr of paraffinic oil without oil bleed. However, the oil gel of the wrist restpresumably contains low molecular weight SBS extended with perhaps 200phr or more paraffinic oil. The SBS then increases oil viscosity, and oil bleedfrom the gel is prevented by encapsulation of the oil in an oil- and abrasion-resistant polyurethane cover. The oil-extended lowmolecular weight SBS canbe readily processed (poured into amold in the wrist rest application) at about150jCbecause of its lower (than SEBS)TODT.Moreover, the slightmiscibilityof the low molecular weight PS end blocks with paraffinic oil would furtherreduce hard and soft phase incompatibility, thereby improving gel process-ability. The high damping characteristics (103) of this gel (perhaps due to thelarge interphase volume created by the low molecular weight of the polymerand the mixing of small quantities of paraffinic oil into the styrene micro-domains) may not be important in the wrist rest application. The lower cost ofSBS compared to SEBS and the high oil loading (which also lowers cost)allowable without bleed due to the oil-resistant polyurethane cover make SBScompetitive in this low-end application where the product UV or thermo-oxidative stability requirement is minimal (104,105). Moreover, intellectualproperty concerning SEBS gels and the large number of SBS manufacturerscompared to SEBSmanufacturers also allow the entry of SBS oil gels into thismarket.
C. SEBS Compound Upper Service TemperatureImprovement
Even though theTg of polystyrene is about 95jC, under stress the polystyrenesegments will flow at a temperature lower than its Tg. SBS loses most of itsstrength at 60–70jC (106,107). In this case, the additional barrier to flow dueto phase incompatibility between the rubber and the plastic is not sufficient toallow the polystyrene microdomains to be good enough anchors to preventdisentanglement of the rubber chains and thus prevent viscous flow. Viscousflow occurs in low molecular weight SEBS (Kraton G1652, Table 4) at
65jC (107), in spite of the increased incompatibility between the rubber andplastic phases compared with SBS (elevated temperature stress–strain data).In SEBS compounds, the presence of PP helps to improve elastic recovery atelevated temperature. Nevertheless, because of the permanent plastic defor-mation of the styrenic domains and their reorganization as discussed earlier,the continuous use temperature of compounds containing high molecularweight SEBS is limited to 70jC, with a 100jC use temperature possible inapplications where there is a limited load on the product.
Table 4 lists the physical properties of paraffinic oil blends of SEBSproducts KratonR G1651, G1650, and G1652 prepared by mixing in alaboratory Brabender. The SEBS materials have approximately the samePS content and are listed in order of decreasing molecular weight. Note thatfor SEBS triblock copolymers, reducing PS molecular weight while keepingthe same weight percent of PS would require a reduction in rubber molecularweight. The compression set increase with decreased SEBS molecular weightcan be related to permanent deformation of the PS microdomains and to the
Table 4 SEBS Block Copolymers: Characterization and Properties
reorganization of these discrete PS domains by interdomain movement of thePS chain ends through the continuous rubber phase. Permanent plastic de-formation of the PS phase may contribute only minimally to the compressionset, because the volume of this phase is only 13% for a 30 wt% PS SEBS,assuming that all the added oil is present in the rubber phase and the inter-phase is neglected (PS density 1.05 g/mL; plasticized rubber density 0.86 g/mL). Hence, lowered SEBS molecular weight must facilitate increasedinterphase movement at the molecular level, as discussed, due to the increasein phase compatibility. Increased phase compatibility is reflected in increasedphase mixing for the lower molecular weight SEBS (higher hardness andhigher M100; M100=modulus at 100% elongation) in comparison with thehigher molecular weight materials (lower hardness and lower M100). Notethat a continuous rubber phase is expected for these SEBS compositions. Theincreased Tg expected for the rubber phase of the low molecular weight SEBSwould contribute to the increased set observed, but it is believed that the bulkof the set observed is due to interdomain movement of the PS segments. Theproperty changes observed when high molecular weight SEBS is processed atdifferent temperatures reflect the difficulty in achieving an equilibriummorphology even after long processing times (compare columns 1 and 2 inTable 4).
The elastic recovery of SEBS compounds at elevated temperature can beimproved by increasing the hard phase Tg while maintaining or exceeding theincompatibility between the rubber and plastic phases over that of SEBS. TheTg of the hard phase can be increased by chemical modification of the PS endblocks in SEBS, by synthesis of triblock copolymers where the PS blocks arereplaced by higher Tg hard blocks, and by compounding with a high Tg
polymer that is miscible with the PS domains of SEBS. Alkylation of thepolystyrene phase of SEBS increases the hard phase Tg, but reduces thecompatibility difference between the rubber and hard phases (93).
A recent publication (107) reviewed the methodology to enhance thehigh-temperature properties of SEBS by chemical modification, whichincreases the Tg of the PS glassy phase.
Poly(a-methylstyrene) (PaMS) has a Tg of 165jC and a-methylstyrene(a-MS) can be polymerized by anionic, cationic, and free radical polymeriza-tion.TriblockcopolymerswithapolyisoprenemidblockandPaMSendblockshave been produced by anionic polymerization, although hurdles have to beovercome due to the low ceiling temperature of aMS (108). An unsuccessfulattempt to synthesize by cationic polymerization an aMS/isobutylene/aMStriblock copolymer has been reported (109). TPEs based on PaMS are notexpected to be of commercial value because of reversion of the polymer tomo-nomer at elevated temperature (110). Hence the use of PaMS hard segmentsis unsuitable for improving the high-temperature compression set of SBCs.
Poly(2,6-dimethyl-1,4-phenylene oxide) (PPO) is the high Tg additive ofchoice for increasing SEBS upper service temperature. Paul and coworkers(111) showed that the solubility of PPO ismuch greater than that of PS itself inthe PS microdomains of SEBS, owing to the exothermic heat of mixing in thecase of PPO. The PPOmolecular weight should be less than or equal to that ofthe PS molecular weight of the SEBS microdomain for miscibility, owing tothe limited conformations available to it in this confined geometry (112–114).Also, PPO of greater molecular weight would lose much of its configurationalentropy and gain only a small amount of translational entropy upon mixing(115). The exothermic heat of mixing can partially compensate for theunfavorable entropic effects associated with PPO confinement in the case ofPPO–SEBS mixing.
Baetzold and Korberstein (96) studied solvent-blended low molecularweight PPOs [Mn/Tg (jC): 1000/116, 2000/161, 6000/196] with low molecularweight SEBS PS [(31.8 wt%), MW = 3400, PEB: MW = 14,600] [PEB =poly(ethylene/butene) SEBS rubber midblock]. PPO is thought to be homo-geneously distributed among the styrene microdomains but heterogeneouslysolubilized within them. PPO is concentrated in the center of the PS micro-domain, with the PPO concentration becoming more diffuse with increasingmolecular weight. The SEBS thusmodified exhibited two high temperatureTg
values—one due to PS and one due to the PPO/PS core of the PS micro-domains. PPO ofMn 2000 could be solubilized into the SEBS PS domains toonly about 26 wt% of the total glassy phase. Beyond this concentration, PPOformed a separate phase. Therefore, there is a limit to which the SEBS hardphase Tg can be increased by compounding with PPO.
An additional disadvantage of this method is the continued presence ofunmixed PS. Koberstein and Baetzold could increase the TODT of SEBS from142jC to 180jC by solution blending with PPO. It should be mentioned thatPaul and coworkers had previously published similar results (102,111). Pauldemonstrated that PPO with Mn values of 15,500–29,400 is miscible in allproportions with high molecular weight SBS (Kraton D1101: 28.8 wt% PS,MW = 14,500; PBD: MW = 67,500; 16 wt% diblock content) and highmolecular weight SEBS (Kraton G1651). The morphology in these materialsis expected to change from a discrete plastic phase to a continuous one withgreater additions of PPO.
Thermoplastic elastomers with improved elevated temperature com-pression set have been produced by melt blending SEBS, PPO, PP, andparaffinic oil (116). Gel compositions with softening points above 100jCwereachieved by solution blending of PPO and SEBS and subsequent plasticiza-tion of the product isolatedwith paraffinic oil (117). Replacement of paraffinicoil in this system with very low molecular weight EPR as plasticizer (toprevent evaporative losses) results in soft SEBS gel compositions (hard-ness<30 Shore A) with good elastic recovery at 150jC (118).
Low molecular weight SEBS is readily processable but has a limitedupper service temperature (see earlier discussion). High molecular weightSEBSor SEBSwith a highTODT has an increased upper service temperature atthe expense of reduced processability. PP that is useful in SEBS compoundingbecause of its partialmeltmiscibilitywith oil and PEB represents another limitbeyond which the upper service temperature of SEBS cannot be raised. PEBis not compatible with the commercially available, higher melting (Tm =240jC), isotactic poly(4-methyl-1-pentene). Hence, the latter plastic cannotbe used to improve the upper service temperature of SEBS compounds.
The increased Tg of the hard phase modified SEBS would necessitate anincrease in compoundmanufacturing temperature, time, andmixing intensityto achieve a near-equilibrium polymermorphology that would be stable at theTPE service temperature. Polymer thermo-oxidative and mechanical degra-dation may preclude these aggressive manufacturing conditions.
With the additional expense of material and compounding, SEBScompounds (presumably modified with PPO) have matched the 70jC tem-perature elastic recovery of PP/EPDM TPVs (Table 5) (119).
Commercially important SBCs include the SBS and SIS polymers (forexample, Kraton Dk) and SEBS and SEPS (Kraton Gk). HydrogenatedSBCs show improved UV and ozone resistance and better strength at higher
Table 5 Property Comparison of SEBS Compounds vs. PP/EPDM TPVs
Manufacturer, trade name
Property
Multibase, Inc., lowcompression set
SEBS compound
Advancedelastomer systems,
PP/EPDM TPV
Multiflex TPE A5001 E LC
SantopreneR101–55W185
Hardness (Shore A) 46 55Tensile strength (psi) 800 at yield 640
Ultimate elongation (%) 600 330Compression set (%)(70jC, 22 hr,ASTM D 395B,constant deflection)
temperatures than the corresponding unsaturated copolymer. Increasingstyrene content increases the strength of the materials; conversely, lowerstyrene content increases elongation of SBCs.
Properties of SBCs include high elasticity and tensile strength, lowdensity, low permeability, good optical clarity and surface appearance, andchemical resistance to acids, bases, and aqueous media. The materials arenormally custom compounded depending upon the application and can beformulated to yield a wide range of performance characteristics. Combina-tions of SBCs with other materials—oils, resins, fillers, processing aids,antioxidants, etc.—yields the desired combination of performance and costdepending upon the application. Compounds usually contain SBC for elasticproperties, plasticizers for softening and improved fabrication, and sufficientthermoplastic resin to reach the desired hardness. Various fillers, includingcarbon black, clay, and talc, lower cost; color is incorporated through directpigment or pigmentmasterbatch addition. Additives include zinc stearate as aprocess aid and stabilization packages for improved temperature or weath-ering resistance.
As already described, SBCs are well known for their oil-holding cap-acity, and relatively large amounts of oil can be incorporated without adetrimental effect on performance. Paraffinic oils are preferred for SBCsbased on their elastic phase compatibility; hardness is decreased, but materialstrength is relatively unaffected. Aromatic oils should be avoided because theywill soften the styrenic hard domain. Compounding with polypropylene iscritical for good injection molding or extrusion fabrication processability forthe SEBS materials because it creates a distinct continuous polypropylenephase.
Compounding of SBCs can be achieved with batch or continuousmixers. Some ingredients, including color or additive concentrates, may beprecompounded; others are added in separate feeders. The final form isusually a finished free-flowing pellet; however, continuous processes formaterial compounding and extrusion have been developed.
Compounded SBCs can be fabricated by conventional thermoplasticmethods including injection molding, blow molding, and extrusion. Thestyrenic character dominates the melt behavior; materials exhibit strongthixotropic or shear thinning behavior at melt temperatures and higherviscosity at higher styrene block Mw. Generally, the melt viscosities are inthe order SEBS>SBS>SIS (due to increasing phase compatibility); onlythe SIS and SBS show a Newtonian region at the low end of the shear raterange.
Numerous precompounded grades have been developed for specificpurposes. After priming, the parts can be coated with paints that are alsoflexible. Applications cover a wide range of products and are found commer-
cially in automotive, medical, footwear, wire and cable, and consumer andindustrial goods. SBCs can be formulated into hot melt adhesives for use inlabels and tapes, eliminating the solvents used in conventional polymersolution products (120). Some of these formulations can be covalentlycross-linked by radiation after coating (120). Elastic films and sheets are usedin medical and diaper films.
Acoustic barriers for dash panels, wheel wells, firewalls, and floorsprovide a significant part of SEBS compound use in automotive applicationsin North America. Other automotive uses include seals, gaskets, airbag doorcovers, and soft-touch interior parts.
In Europe SBS block copolymers are widely used as asphalt modifiers(121). Used at relatively low concentrations, these materials provide recycla-ble and safe solutions to improve the performance of asphalt by forming athree-dimensional structure within the material. Special polymers have beenengineered to provide the appropriate balance of compatibility and flowproperties for road structures and roofing applications (121).
VII. THERMOPLASTIC VULCANIZATES
A significant advance in polyolefin-based thermoplastic elastomers resultedfrom the discovery that EPDM rubber, when selectively cross-linked undershear (dynamic vulcanization) duringmelt blendingwith a compatible plastic,namely isotactic homopolypropylene, results in a thermoplastic elastomerwith mechanical properties and fabricability far superior to those obtainedfrom a simple blend of the elastic and plastic materials (122–129). Indeed, theperformance, price, and environmental impact of PP/EPDM TPVs haveprovided impetus for replacement of thermoset rubber by these thermoplasticelastomers. Penetration of the thermoset rubber market by PP/EPDM TPVshas been made possible by the breadth of the product service temperature(�40jC to 135jC), hardness range (35 Shore A to 50 Shore D), excellentfabricability and fabrication economics, and the ability of product scrap to bereprocessed, among other desirable environmental characteristics. PP/EPDMTPVs are products of commerce in thermoset rubber replacements throughfinished part cost savings realized by fabrication, design, and materialeconomics. Compared to SEBS compounds, PP/EPDM TPVs exhibit betterelastic recovery at a higher service temperature (100jC vs. 70jC). In ‘‘static’’applications PP/EPDMTPVs can provide service at 135jC, versus 100jC forSEBS compounds. Very soft TPEs (0–5 Shore A) are based on SEBS; PP/EPDM TPVs with hardness lower than 35 Shore A are not commerciallyavailable.
Dynamic vulcanization is the process of producing a thermoplastic elastomerby selective cross-linking of the rubber phase during mixing of a technolog-ically compatible or compatibilized rubber and plastic blend of high rubbercontent while minimally affecting the plastic phase. Rubber cross-linking isaccomplished only after a well-mixed molten polymer blend is formed, andintensive blend mixing is continued during the curing process. The elasto-meric thermoplastic vulcanizate thus formed should ideally consist of a plasticmatrix that is filled with 1–5 Am cross-linked rubber particles.
B. Development of Dynamic Vulcanizates:Historical Perspective
Dynamic vulcanization has its origin in the work of Gessler and Haslett (130)at Esso, where they demonstrated that carbon black–filled blends of isotacticPP and chlorobutyl rubber with good tensile strength could be obtained bycuring the rubber (with zinc oxide) after the components were blended on amill at room temperature, by further milling the blend at the curing temper-ature and above the melting point of PP. Tensile strength was lower when theblend was ‘‘statically’’ cured. Both ‘‘dynamic’’ and ‘‘static’’ cure resulted inthermoplastic compositions. The first patent claim, limited to chlorobutylrubber, included blends with up to 50 wt% rubber in PP that could be curedwith any curative that did not break down the PP plastic material. Theincorporation of plasticizer oils into the blends is one of the items outlined inthe second claim. The goal of this work may have been the impact modifica-tion of PP, because the rubber content of the blend was limited to 50 wt% inthe patent claims. Captured in this work were the essential attributes ofdynamic vulcanization as practiced today, including the use of rubberplasticizer oils, except for recognition of the importance of curing the rubberphase only after the formation of an intimate plastic and rubber blend and thevalue of compositions containing a high rubber content.
Gessler and Haslett did not continue their research on the dynamicvulcanization of rubber and plastic blends; the work of Fischer representedthe next advance in this technology. It was shown that the properties ofpolyolefin blends with EPM or EPDM rubber (thermoplastic olefins; TPOs)could be dramatically improved if the rubber was first either statically ordynamically (on a mill, Banbury, or extruder) cured to a gel content of up to90% before being melt blended with the plastic. The gelled rubber was stillprocessable (could be ‘‘banded’’ on a mill) prior to melt blending with theplastic. The TPO rubber content could be as high as 90%. The improvementin TPO properties was quantified by an increased ‘‘performance factor’’
(tensile strength (psi) � elongation at break (%) divided by elongation set atbreak) (131,132). Another route to TPOs with improved properties wasachieved by the use of EPM or EPDM rubber that was branched in acontrolled manner during rubber production (133).
Fischer’s work on TPOs culminated in the dynamic vulcanization (in aBanbury) of molten blends of PP/EPM or PP/EPDMwith peroxide (134). Tomaintain thermoplastic processability (‘‘banding’’ of the final product on amill or extrudability as a measure of product processability), the rubber curestate had to be limited. In the patent, the maximum cure state claimed for therubber is 90% gel. Dynamic vulcanization tremendously increases meltviscosity over that of the TPO melt, which increases with increased rubbercure state, thus reducing thermoplastic vulcanizate (TPV) processability. TPVphysical properties, however, improve with increased rubber cure state.Increased tensile strength, improved compression and tension set, lower swellin hydrocarbon oils, and improved flex fatigue and abrasion resistance aremanifested in a TPV in comparison with the corresponding TPO. Neverthe-less, because of PP plastic breakdown by the peroxide rubber curativeemployed by Fischer, the TPVs of the illustrative patent examples did notachieve their full property potential. To improve TPV melt processability,Fischer indicates the use of very limited quantities (fone part per 100 parts ofrubber) of process aids (e.g., epoxidized soybean oil, polymeric slip aids).Uniroyal’s polyolefin thermoplastic rubber (TPRR, commercialized in 1972)is based on Fischer’s work on dynamic vulcanization.
About the time Fischer was pursuing his studies on dynamic vulcani-zation, Paul Hartman at Allied Chemical Corporation claimed that butylrubber could be grafted onto polyethylene by using a difunctional resole-typephenolic resin as rubber curative while the rubber and plastic were meltblended on a mill. The grafting was thought to occur via the end olefinicfunctionality in PE. The grafted material exhibited superior physical proper-ties and had a greater capacity to disperse fillers such as carbon black and talcthan the simple melt blended product. Rubber cross-linking was avoided byusing judicious amounts of the low functional phenolic resin curative. Graftedproducts of butyl rubber, EPDM, or diene rubbers such as SBR and NBRonto PP, PE, or poly(1-butene) were claimed (135–137). The completesolubility of the products of this invention in hot xylene was taken as proofof grafting of the rubber onto plastic and the absence of cross-linked rubber.In all probability, the expected grafting reaction was minimal, and the rubbersimply underwent chain extension in the presence of limited amounts ofcurative during dynamic vulcanization.
Monsanto entered the field of dynamic vulcanization with a patent byCoran et al. (138) that extended the work of Fischer to dynamic vulcanizationof diene rubbers in a polyolefin matrix. The inventors demonstrated that
thermoplastic compositions could be obtained even when the rubber wascured to a high cure state as opposed to Fischer’s finding that the gel contentof the rubber obtained in dynamic vulcanization should be less than 100% inorder to maintain product thermoplastic processability. In addition, theMonsanto inventors realized that TPV physical properties improve whendynamic vulcanization is continued to achieve a high rubber cure state. TPVphysical property improvement was also attributed to the presence of smallrubber particles (less than 50 Am in diameter) dispersed in a plastic matrix.Subsequently, high rubber content TPVs based on butyl rubber were claimedbyMonsanto (139). The patent claims included compositions containing highlevels of plasticizer oils.
In a coup de grace (140), Coran et al. demonstrated that, contrary to theFischer partial cure requirement, PP/EPDM TPVs with both excellentphysical properties and processability can be obtained when the rubber iscured to a high cure state. Sulfur, which does not degrade PP, was the curativeof choice in the patent examples. Processable TPVs with a high rubber curestate could also be obtained by peroxide cure in the presence of a bismaleimidecoagent that undoubtedly limited PP breakdown in addition to allowing theachievement of a fully cured rubber phase. It was also recognized that thepresence of a high level of paraffinic oil allowed the preparation of soft,processable TPVs with excellent elastic recovery. On TPV plastic phasemelting, the oil in the rubber could partially partition into the molten plasticand also form a separate oil phase (nonequilibrium conditions probably existdue to the short TPV processing time). These factors result in a considerableimprovement in TPV processability.When the TPVmelt is cooled, the free oiland the oil rejected from the crystallizing plastic are reabsorbed into therubber and the amorphous plastic domain. At the same time, Gessler andKresge (141) disclosed that PP/EPDM or EPM TPOs that were producedwith high molecular weight rubber had desirable physical properties but werenot processable. The TPOs exhibited both good physical properties andprocessability if paraffinic oil was added to the compositions.
Monsanto began its effort to commercialize TPVs with PP/EPDM/paraffinic oil compositions that were cured with sulfur. TPV morphologyconsisted of a PP matrix that was filled with cross-linked, micrometer-sized(5–15 Am) rubber particles. TPVmechanical properties and fabricability weredependent upon rubber particle size, with the size just indicated beingpreferred. During TPV processing, however, the rubber particles increasedin size, presumably due to the breakage and re-formation of the polysulfidiccross-links that occur in the melt during particle collision. This unstable meltmorphology (‘‘melt stagnation’’ or phase growth of the dispersed rubber)resulted in poor and variable product fabricability and mechanical properties
(142). Evolution of gases with an unpleasant odor during manufacturing isanother serious drawback of sulfur-cured TPVs.
This final obstacle to PP/EPDMTPV commercialization was overcomeby Abdon-Sabet and Fath (143) by the use of a resole-type phenolic resin ascurative.Melt stagnationwas avoided by the formation of thermo-oxidativelystable cross-links in the rubber particle. Improved TPV resistance to hydro-carbon oils and compression set also resulted from the use of the phenolicresin curative. The advantages of using the phenolic resin in diene rubber–based TPVs were simultaneously recognized by Coran and Patel (144).
The granting of the Monsanto TPV patents led to heavy but unsuccess-ful opposition by several corporations. Uniroyal had entered the TPVmarketwith a product for which the rubber phase was partially cured, butSantopreneR rubber by Monsanto had superior physical properties due toits fully cured rubber phase that was achieved without plastic phase break-down. Leading TPV suppliers today include ExxonMobil/Advanced Elasto-mer Systems (SantopreneR), Mitsui (MilastomerR), Sumitomo (SumitomoTPE), and DSM Copolymer (SarlinkR).
C. Principles of Dynamic Vulcanization
Thermoplastic vulcanizates are complex systems that, when formulated andprocessed correctly, result in materials that show significant fabricationadvantages over thermoset rubber. Six key requirements have been identifiedin the preparation of polyolefinic TPVs:
Principle I. Rubber and Plastic Compatibility
The first principle of dynamic vulcanization is that the rubber and plasticshould be very compatible but not melt-miscible. If the plastic were misciblewith the rubber, cross-linking of the rubber would result in the inclusion oflarge portions of plastic into the rubber particles, reducing the amount of
Principles of dynamic vulcanizationI. Rubber and plastic compatibilityII. Interphase structure
III. Plastic phase crystallinityIV. Rubber vulcanizationV. Morphology control
plastic in the continuous phase. The resultant product would lose thermo-plastic processability; the product of dynamic vulcanization would be apowder. Moreover, because the plastic’s Tg is usually much higher than thatof rubber, if the plastic is miscible with the rubber over a broad temperaturerange the blend would not be elastomeric owing to a high average Tg. Theextent of compatibility between the rubber and plastic required for TPVs withgood physical properties and processability is difficult to quantify. Evenbetween polymers that are considered technologically incompatible, such aspolystyrene and poly(methyl methacrylate), an interphase thickness of 50 Ahas been measured (145). Presumably, the high entropic penalty for demixingresults in entanglement between different polymers in the interphase ofincompatible polymer blends.
The more compatible the polymers, the greater the interphase thicknessand interpolymer entanglements, and the better the mechanical properties ofthe polymer blend (146). The difference between the ‘‘critical surface tensionfor wetting’’ of polymers is considered to be a rough measure of polymerinterfacial tension (147,148). Based on surface tension values, PP/EPDMblends (EPDM not characterized) are among the most compatible of therubber and plastic blends evaluated by Coran and Patel in dynamic vulcani-zation (147).
The excellent compatibility of PP and EPR is reflected in the fine blendmorphology that can be generated by melt mixing of the components(141,149–153). Indeed, EPR is the impact modifier of choice for PP becauseof the good interfacial adhesion between these components (154,155). Datta,and Lohse (156) showed that in 80:20 by weight PP/EPR blends, EPR rubberparticle size could be further reduced by the addition of an iPP-g-EPcopolymer referred to as a compatibilizer. There should be a distinctiondrawn between emulsification and compatibilization. A diblock polymeracting as a true compatibilizer should have its block segments anchored inthe phases being compatibilized. This would also imply an increase ininterphase thickness over the uncompatibilized blend, which would alsoresult in improved mechanical properties for the blend. A true compatibilizeris more likely to be formed by the in situ reaction of components that arealready present in the phases being compatibilized. An example would be theimpact modification of nylon by the melt blending of this plastic with amiscible blend of EPR and maleated EPR to generate fine particulate EPR ina nylon matrix (157). The amine end groups of the nylon would then reactwith the grafted maleic anhydride moiety in the EPR to generate in situ ablock copolymer in which the blocks are truly anchored in the phases beingcompatibilized.
In emulsification, as in the emulsification of oil in water by soap, theemulsifier is adsorbed only onto the surface of the particulate dispersed phase
(158) and may also be present as a separate phase that is an additional barrierin preventing coalescence of the dispersed phase. These phenomena have beenobserved in the case where block copolymers have been melt mixed withincompatible polymer blends (159,160). There is a recent report of theinability of PP-b-EPR to function as a compatibilizer in PP/EPR blendsdue to cocrystallization of the PP segment of the compatibilizer with the PPphase and rejection of the EP block into the interlamellar space of the PPcrystals (161).
Key aspects pertaining to the interfacial adhesion of immiscible poly-mers are briefly reviewed by Adedeji and Jamieson (162), who studied particlesize reduction due to emulsification and compatibilization in solution blendsof styrene/acrylonitrile (SAN) copolymers and polystyrene (PS) producedwith and without added poly(methyl methacrylate)-block-polystyrene(PMMA-b-PS). Compatibilization or emulsification of the dispersed PSphase in SAN depended upon the polymer molecular masses, including thePMMA-b-PS compatibilizer segmental molecular mass. These determina-tions were made by studying crazes generated by the mechanical deformationof blended polymer films and observing preserved (due to emulsification) orfractured PS domains caused by efficient stress transfer across the SAN/PSinterphase (due to compatibilization).
Chun and Han (94) also pointed out the importance of distinguishingbetween emulsification and compatibilization of polymer blends. The previ-ously mentioned (156,163) observation of improvedmechanical properties onsize reduction of the dispersed EPR phase in the iPP matrix, by means of theiPP-g-EPR additive, may not be due to improved interfacial adhesion. Themechanical property improvement in impact-modified (by polyolefin rubber)isotactic polypropylene has been shown to be dependent upon rubbermolecular weight (151,164–166), microstructure (151), and composition(151,165,167,168), rubber particle size and particle size distribution (151,164,169–177), the ability of the rubber particle to cavitate (165,171,173,174,178–180), optimal interparticle distance (164,181–184) (and thereforePP ligament thickness), and perhaps modification of the PP crystal phase bythe rubber particles (152,176,180,185,186), a compatibilizer (187), or PPnucleating additives (188,189).
Many of the aspects discussed in connection with the mechanicalproperty improvement in impact-modified iPP are interrelated. For example,the rubber particle size obtained in an iPP/polyolefin rubber blend woulddepend upon rubber melt viscosity (and hence rubber molecular weight) andprocessing conditions. Adhesion of the rubber to the iPP matrix woulddepend upon rubber molecular weight, which can be related to the extent ofmiscibility of the low molecular weight ends of the two components (190).However, if the rubber phase is cross-linked—for example, to stabilize the
melt morphology—interphase adhesion would be affected. The ability of arubber particle to cavitate would depend upon rubber molecular weight andthe extent of any cross-linking present. The iPP crystal structure obtained oncooling iPP/polyolefin blends would also depend upon rubber particle sizeand cure state.
The impact strength of plastics would depend upon the morphology-dependent deformation behavior (shear yielding, crazing, and rubber particlecavitation) that ensues under the impact conditions. Impact strength is alsodependent upon processing conditions (191,192), which would affect rubberparticle size and shape and cause built-in stresses due to ‘‘frozen in’’ rubberparticle shape and differential shrinkage between the plastic and rubber. Dur-ing the cooling ofmolten impactmodified (with particulate polyolefin rubber)isotactive polypropylene, for example, any voids created by differentialshrinkage between the rubber and plastic phase will be filled in as long asthe phases are mobile. Although there is considerable shrinkage in the plasticphase from the melting point to the crystallization temperature, once themorphology is frozen (corresponding to the crystallization temperature), therubber particles shrink more than the plastic on cooling to room temperature.This differential shrinkage would have caused the rubber particles to shrinkaway from the plastic/rubber interface were it not for the good adhesion be-tween the phases, but imposes triaxial tension on the rubber phase (176,193).
On bending, a piece of impact-modified polypropylene sometimesresults in stress whitening (‘‘blush’’) due to scattering of light at the creaseline, presumably owing to rubber particle cavitation. Polypropylene ‘‘blush’’can be reduced by compounding in polyethylene, which is more compatiblewith the rubber than plastic, and hence resides within the rubber particles.This allows a reduction in the rubber amount and hence limits rubbershrinkage and changes the stress distribution across the rubber particles.Rubber particle cavitation is prevented due to the excellent adhesion of therubber to both the polypropylene and polyethylene phases. The impactproperties of the system are then also improved while minimizing stiffnessdecrease (194). Fibrillar rubbermorphology is undesirable in impact-modifiedplastics, because these types of rubber particles presumably act as inefficientcraze initiators (166,195). Finally, the behavior of plastics on impact woulddepend upon temperature and test speed, and at high test speeds adiabaticheating in the deformation zone has to be taken into account (196,197).
A review of the structure and properties of polypropylene/elastomerblends has been published (198). Modification of iPP with particulate rubberoffers improvedmaterial impact strength at reduced stiffness. It is desirable toachieve good material impact properties while maintaining stiffness. High-impact polystyrene (HIPS) is filled with appropriately sized rubber particles,which allows good impact properties, but these rubber particles containtrapped particulate polystyrene (‘‘salami’’ structure), resulting in the material
having a reduced rubber content, which limits the stiffness decrease on impactmodification of polystyrene (199). This technique has also been applied tobalancing iPP rigidity with toughness (200,201).
Compatibility and miscibility in PP/EPDM molten blends would bedependent upon polymer molecular weight, the ethylene/propylene (E/P)ratio and E/P compositional distribution in the rubber molecules, and thelength of sequences of the structural units in the rubber. Although iPP andatactic PP aremiscible in themelt (202,203), iPP and sPP are notmelt-misciblein all proportions, and hence the tacticity of iPP may have an influence on thecompatibility and miscibility in PP/EPDM molten blends. Blends of PP andEPR copolymers have been shown by neutron scattering to be immiscible inthe melt, even when the ethylene content of the EP copolymer is as low as8 wt% (203). Kyu and coworkers (204) found that iPP with anMn of 247,000and EPDM (70% ethylene, 5% ENB,ML 1+ 4 at 125jC= 55) are misciblebelow the melting point but above the crystallization temperature of iPP (anlower critical solution temperature (LCST) was observed) (204). Somemiscibility between iPP and EPDM is indicated in the solid state because ofa Tg increase of the rubber in the melt blended product (192). In PP/ethylene-1-octene plastomer blends, the lower rubber Tg observed has been attributedto trapped stresses in the rubber phase caused by a mismatch in the thermalcontraction characteristics of the plastic and rubber (193).
It is well established that iPP/EPDM blends are immiscible at temper-atures well above the iPP melting point at which dynamic vulcanization isconducted. The partitioning of paraffinic oil between molten PP and EPDMdoes not change the system’s immiscibility characteristics (86).
Table 6 lists the formulations and properties of PP/EPDM/paraffinicoil melt-blended products (TPOs) and the corresponding TPVs. Paraffinic oilis completely miscible with EPDM and molten PP. The formulations weremixed in a laboratory Brabender at 180jC followed by curing of the moltenblend under shear for TPV preparations and subsequently compressionmolded at 200jC to produce plaques for testing. The morphology of theproducts as probed by atomic force microscopy (AFM) is presented in Figure5, with product numbers as in Table 6. Rubber is represented by the darkareas, and plastic domains by the lighter color. Products 1–4 have acontinuous rubber phase and a discrete plastic phase, whereas the phasemorphology is reversed in products 5 and 6. Oil was extracted from the TPVsamples by Soxhlet using the cyclohexane/acetone azeotrope, followingwhichthe dried samples were subjected to extraction by cyclohexane at roomtemperature (24 hr) to determine the amount of soluble rubber present. Thedata indicate that all the TPV samples had a high level of cross-linked rubber.The volume percent rubber in the formulations was calculated assuming thatall the oil is in the rubber phase, with none in the amorphous portion of theplastic phase, which is not quite correct (205).
Despite the high rubber cure state and rubber continuous/discreteplastic solid-state morphology, products 2 and 4 are still thermoplastic.Perhaps in the melt state the rubber and plastic phases are cocontinuous orthere is a continuous plastic phase. In both cases, thermoplastic processabilityis expected. For the TPVs in question, particulate rubber (of undeterminedsize) could form a continuous phase by impingement and entanglement ofpolymer chains at the points of contact of the rubber particles. The TPVmorphology is dependent upon the level of shear introduced into the meltduring dynamic vulcanization.
Transformation of Thermoplastic Olefins to ThermoplasticVulcanizates
On transformation of TPO to TPV there is a substantial improvement in thedesirable elastomeric physical properties such as elastic recovery (compres-sion set, tension set) and tensile strength (Table 6). This is counterintuitive,because compatibility between the rubber and plastic should decrease oncross-linking the rubber and result in poorer TPV properties compared withthe corresponding TPO. The observed TPV properties can be explained if theparticulate rubber is firmly anchored in the amorphous portion of the plastic
Table 6 Dynamic Vulcanization of PP/EPDM/Oil Blends: Compression-MoldedPlaques from Brabender Preparations
phase. EPDM has a very low molecular weight between entanglements (1660for EP rubber with 46 wt% E) (206), which is not much higher than that ofpolyethylene (1250) (206). The molecular weight between entanglements forPP is 7000 (207). Then, owing to the PP/EPDM compatibility already dis-cussed, the two polymers are highly entangled in the interphase, and theseentanglements are trapped when the plastic crystallizes. It is conceivable thata trapped PP segment on the rubber particle surface is a tie molecule that findsitself anchored to the rubber particle because it is part of one or more PPcrystalline segments. This thinking is not without some precedent. In PS/PMMAblendswhere the interfacial tension is relatively small, no difference inthe mechanical behavior of the interface could be observed in the presence orabsence of PS-b-PMMA copolymers of different molecular weights (162).This has been attributed to mechanically effective entanglements that arealready present in the PS/PMMA blends. Lohse could not improve theproperties of PP/EPDM TPVs (unpublished results) by adding PP-g-EPR(156) as compatibilizer. Coran and Patel (208) could prepare PP/nitrile rubber(NBR) TPVs with good elastic properties only if the plastic was first pre-treated with the phenolic resin curative. The PPwas thought to bemodified bythe phenolic resin (presumably due to end unsaturation in the plastic) suchthat a PP-g-NBR compatibilizer could be formed in situ during dynamicvulcanization. This approach did not improve the properties of PP/EPDMTPVs (unpublished results), which suggests that good adhesion already existsbetween the rubber and plastic phases in this system.
The hardness change on TPO-to-TPV transformation is dependentupon the change in product morphology and crystallinity in the plastic phaseand is also due to cross-linking of the rubber phase. Because the rubber phaseis still continuous when TPO Samples 1 and 3 are converted to TPVs 2 and 4,respectively (Table 6), the change in hardness can be related to cross-linkingof the rubber. Because both the highly plastic-loaded formulations 5 and 6exhibit a plastic continuous morphology, the lack of change in hardness onTPO-to-TPV transformation can be readily rationalized.
Inoue and Suzuki (209,210) studied the impact properties of PP (70wt%) / EPDM (30 wt%) melt blends before and after dynamic vulcanizationwith a curative that is not expected to affect PP (N,N-m-phenylenebismalei-mide/2,2,4-trimethyl-1,2-dihydroquinoline). No change in rubber particlesize results from dynamic vulcanization. The considerable improvement inimpact properties observed after dynamic vulcanization was related toincreased interfacial adhesion between the rubber and the plastic (due tothe in situ formation of a compatibilizer during cure by grafting of rubberonto the plastic), which caused increased shear yielding and crazing in thedamaged plastic zone created by impact testing. It was also noted that thecross-linked EPDM particles acted as nucleating agents and decreased PP
spherulite dimensions. The extent of the improved impact strength observedon dynamic vulcanization due to the change in plastic morphology was notestablished. The un-cross-linked rubber particles in the iPP were shown tocavitate, and energy dissipationwas thought to occur by craze initiation at therubber particle sites subsequent to cavitation, with the cavitation process itselfcontributing to only a modest improvement in impact strength. Rubberparticle cavitation was not observed on impact testing of the modified iPPcontaining cross-linked rubber particles as suggested by Ishikawa et al. (211).Instead, plastic shear yielding was thought to occur at the rubber particlesites, which led to the formation of stronger crazes (compared with the casewhere iPP contained un-cross-linked particulate rubber) due to stronger cold-drawn fibrils spanning the craze. These strong fibrils presumably resultedfrom the in situ rubber-to-plastic grafting that occurred on cross-linking ofthe rubber particles, as mentioned earlier. Similar results have been reportedby Krulis et al. (212) for dynamically vulcanized PP (80 wt%) / EPDM (20wt%) blends (sulfur cure).
Differential shrinkage on cooling of the blends in question would causethe rubber particles to be in triaxial tension (see previous discussion) in bothTPO and TPV, but the built-in strain should be detrimental to impactstrength. The cross-linked rubber particles of the TPV are less likely todissipate energy by cavitation compared to the TPO, which contains gel-freeparticulate rubber. Because iPP is the major component of the impact-modified plastic and impact modification of iPP by cross-linked rubberparticles has a dramatic effect on the plastic crystal structure, it is reasonableto propose that the increase in impact strength observed on TPO-to-TPVtransformation is due to modification of the plastic phase crystal structure bythe cross-linked rubber particles according to the significant advance in ourunderstanding of the impact behavior of plastics due to Argon and coworkers(213–215). This work suggests that when the plastic crystalline structure thatis nucleated around a rubber particle percolates throughout the specimen,impact properties are optimized. Therefore, the plastic ligament thicknessmust be small enough to allow overlap of the crystalline structures that arenucleated around the rubber particles.
Semicrystalline polymers such as nylon-6 and polypropylene that arenormally ductile fail in a brittle manner under impact loading. The impacttoughness of these materials can be improved by the incorporation of rubberparticles that are bonded to the matrix, at the expense of a reduction inmaterial stiffness. The data of Argon and coworkers indicate that there is adramatic increase in material impact toughness when the interparticle matrixligament thickness is below a critical dimension. This phenomenon hasgenerally been attributed to the overlapping of stress fields created by theappropriately spaced rubber particles of the ‘‘right’’ dimensions. Certainmelt-blended PP/calcium carbonate composites are stiffer and have greater
Izod impact strength than the unfilled plastic (215). This effect was notobserved for nylon-6 (214). The key requirement for plastic toughening inthese cases is rubber particle cavitation or debonding of the inorganic particlefrom the matrix prior to initiation of matrix plastic flow. The presence ofrubber particles or particulate inorganic material induces the formation of alayer of orientedmatrix crystals of well-defined thickness around the particlesupon cooling after melt mixing. When the matrix ligament thickness is smallenough to allow these oriented crystal structures to percolate throughout thespecimen, a dramatic increase in material toughness is achieved. Nylon-6could not be toughened with calcium carbonate particles using this concept,owing to only partial debonding of the rigid particles from thematrix, therebycausing stress concentration in the early phases of the impact response thatcaused a reduction in fracture toughness.
Extending this concept to the TPO transformation, it is known that theimpact behavior of materials is controlled by the microstructure-dependentmolecular response that is initiated under the specific impact test conditions(temperature, strain rate, adiabatic heating). In a PP/EPDMTPO system, theimpact response has been demonstrated to proceed from a brittle response atlow rubber volume with cavitation of rubber particles on impact, to a ductileresponse characterized by matrix shear yielding without rubber particlecavitation or debonding of the rubber particles from the matrix at a higherrubber volume fraction. Perhaps the cavitated rubber particles acted as stressconcentrators and promoted matrix crazing in the case of brittle failure (216).Increasing rubber particle volume in the PP matrix from 0% to about 40%resulted in a step increase in impact strength when the rubber volumeconsisting of the ‘‘right’’-sized rubber particles established a PP ligamentthickness of 0.1 Am, presumably due to establishment of the ‘‘percolated’’ PPcrystal structure already discussed. Increasing the rubber particle volumebeyond a certain level in the PP/EPDM compositions resulted in a drop inimpact strength due to the decrease in matrix phase volume. Fracturetoughness decreased with increasing strain rate as expected, but in somecases, on further increase in strain rate, PP/EPDM blend fracture toughnessincreased due to adiabatic heating of the impact damage zone and thesurrounding volume. Also, in the TPO ductile region, fracture energy initiallyincreases, then decreases, as temperature increases due to the decrease inmatrix yield strength with temperature (217). As already discussed, physicalproperty improvement on TPO-to-TPV transformation can be rationalizedby presuming that an improvement in adhesion between the rubber particleand matrix accompanies this transformation. This improved adhesion wouldassist interphase stress transfer during impact loading of the Inoue andSuzuki TPVs (209,210), but the unexpected improvement in TPV impactresistance over that of TPO was probably due to PP crystal structuremodification.
Coran and Patel (127) studied the properties of dynamic vulcanizatesobtained from 99 rubber and plastic combinations. Only a very limitednumber of TPVs of this study were technologically useful, with the properties(physical properties, processability, and cost) of PP/EPDM-based productsfar surpassing those of the other materials. Although a wide variety ofthermoplastics and elastomers could be combined to form TPVs, the bestresults, based on tensile strength and compression set, were attributed to thecompatibility of PP and EPDM.
Principle II. Interphase Structure
The second principle of dynamic vulcanization is that the rubber and/orplastic should have a low molecular weight between entanglements, so thatthe cross-linked rubber particles formed are firmly anchored in the amor-phous portion of the plastic phase by trapped entanglements. The formationof this so-called interphase is critical to the elastic properties of the TPV.
Principle III. Plastic Phase Crystallinity
The third principle of dynamic vulcanization is that the plastic phase shouldbe highly crystalline so as to provide sufficient cross-links (thermoreversible)for good elastic recovery. The plastic crystal melting point should be highenough to provide the desired elastomer upper service temperature. The plas-tic Tg should be low for improved TPV processability and low-temperatureproperties. There should be no plastic phase transitions between Tg and Tm
because that would be detrimental to TPV high-temperature elastic recovery.Note that iPP exhibits an a-crystal transition between 30jC and 80jC (218).TPV processability is enhanced by plastic materials with a broad molecularweight distribution.
The continuous plastic phase controls both the upper service temper-ature and melt processability of the TPV. In the case of iPP/EPDMTPVs, the165jC peak DSC melting point (which is lowered to about 155jC incommercially available 60 Shore A TPVs due to plastic crystal structureconsisting of fragmented lamellae (see Sec. VII.D) and due to the presence ofparaffinic oil (small effect for kinetic reasons)) of the isotactic PP allows a100jC upper service temperature, which can be extended to 135jC in no-loadapplications. In polyolefinic systems, where polar interactions are absent,plastic ‘‘hydrodynamic volume’’ matching with the rubber will allow maxi-mum compatibility with the rubber from the melt to the solid state bymaximizing the entropy gain in the interphase region. For polyolefins, thiscan be related to the Tg of the plastic, which should be as close as possible tothat of the rubber. PP has a Tg of 0–10jC [depending upon rate and type(DSC, DMA) of measurement], which is relatively close, for a plasticmaterial, to the Tg of EPDM (�50jC for a 60 wt% ethylene EPDM with
4.5 wt% copolymerized ethylidenenorbornene as cure site). Polyethylene witha Tg of –80jC is more compatible with the EPDM rubber under consider-ation, but this increased compatibility raises TPV melt viscosity beyond thefabrication capability of conventional plastics-processing equipment (219).The relatively low Tg of iPP allows a low melt viscosity for this plastic,because on melting at 165jC the melt is already 165jC above the Tg. This lowPP melt viscosity is critical to TPV processability, because this viscosity israised considerably on being filled with cross-linked EPDM particles.
It is now readily recognized that a completely amorphous polymer maynot be suitable as the TPV plastic phase from the standpoint of processability.For example, commercially available, highly amorphous, bisphenol A–basedpolycarbonate (PC) with a Tg of 150jC will allow a use temperature wellbelow 150jC. However, the melt viscosity of polymers decreases graduallyabove Tg, and PC is economically processable only just above 300jC. FillingPC with cross-linked rubber particles would raise the melt viscosity such thatthe material would be processable only above the plastic and rubber decom-position temperatures. Therefore, a completely amorphous plastic is unsuit-able for use as the TPV plastic phase component.
Because the plastic material will become part of the TPV elastomericsystem, it is reasonable to choose a ductile, as opposed to a brittle, material forthis purpose. The mode of failure of a plastic material under, say, uniaxialtension will depend upon the temperature, the strain rate, adiabatic heating,the flaws present in the sample, and sample size. Abrittlematerial will fail aftera ‘‘small’’ deformation, and the damage zone around the failure surfacewill be‘‘limited.’’ For brittle material, failure generally starts with the development,perpendicular to the sample tensile direction, of microcracks (crazes) that arespanned by load-bearing cold-drawn fibrils. The crazes rapidly coalesce into acrack that rapidly propagates perpendicular to the tensile direction, leading tomaterial failure. ‘‘Crystal’’ polystyrene (PS, so called because of its hightransparency) is a high Tg (f100jC), completely amorphous brittle materialwhose failure in tension begins with the formation of crazes that can be 0.1–2Am thick and 50–1000 Am long with fibril diameters varying from 4 to 10 nm(220). Preceding craze formulation there is an expansion in sample volume dueto an increase in polymer hydrodynamic volume as the system attempts toreduce the applied stress. Argon and Hannoosh (221) showed that for small,highly perfect samples of PS, deformation by shear yielding precedes crazing.Sample preparation is important, because dust particles have been observed toact as craze nuclei in a thin film of PS, as observed by transmission electronmicroscopy. Impact-modified PS has been shown to deform by simultaneousrubber particle cavitation, crazing, and shear yielding (199).
If a bar of material is placed in uniaxial tension, the principal tensilestress acts on the area of the bar that is perpendicular to the tensile direction.If the material is isotropic, the maximum shear (or ‘‘sidewise’’) force acts on
an area inclined at 45j to the tensile axis (222).Macromolecules will align andflow past each other (shear yielding or plastic flow) wherever the stress in theslip plane exceeds the material yield stress. Deformation is initiated at a site ofstress concentration created by a structural imperfection in the material,which may be caused by energetically unfavorable molecular arrangements(‘‘built-in’’ stresses), voids, or the presence of foreign matter. The stressconcentration is created by the ‘‘bending’’ of the stress field around theimperfection. It is preferable, of course, that the plastic material elongate tohigh strain by shear yielding rather than crazing before rupture, if it is to bechosen as the TPV plastic phase.
Once a damage zone is created around a site of stress concentration,more material is drawn from the sides toward the damage zone as the testspecimen ‘‘necks’’ in the tensile direction. If the stress field generated is such
Figure 6 Plane stress condition (stress in yz plane only). Sample width y is large incomparison with the damage zone that is centered at the crossing of the axes andcreated by a tensile force in the z direction. Material contained in the width y restricts
reduction in this dimension, generating stresses pointing away from the center in they directions in response to the tensile force in the z direction. Sample thickness in thex direction is much smaller than the width, and the material from the x direction
feeds the sample neck as it elongates in the z direction (low stress in the x direction).Note that since the sample dimension will decrease minimally in the y direction,plane strain conditions prevail in the xz plane.
that material from the sides cannot be drawn in fast enough (depending upontest conditions and molecular characteristics), polymer chains may slip pasteach other or rupture under load in the damage zone, creating a void that cancause an increase in stresses across the now reduced area of the test specimen,leading to brittle fracture. If the thickness of the specimen is small incomparison to the damage zone (plane stress conditions, Fig. 6), one canexpect shear deformation, and in the opposite case (plane strain conditions,Fig. 7) for the same applied strain rate, brittle failure. As the strain rate isincreased, the failure mode of a material can change from ductile to brittle. At
Figure 7 Plane strain condition (strain in the xz and yz planes). Sample width yand thickness x is large in comparison with the damage zone that is centered at the
crossing of the axes, and created by a tensile force in the z direction. Reduction ofdimension in the xy plane is restricted due to large sample width and thickness, thuscreating stresses pointing away from the center in the xy plane, resulting in a triaxialor dialational stress field. Most of the strain occurs in the z direction, creating a large
stress concentration in the damage zone that is unable to draw in much material fromthe xy plane. Chain breakage and/or slippage of the macromolecules in the damagezone leads to transfer of stress to other sites of stress concentration in the sample,
and the test specimen breaks in a brittle manner with a small strain in the z direction.
very low strain rates, a material may show brittle behavior, as in the stresscracking of polyethylene (223). A material is considered to be brittle if, underthe test conditions in question, failure occurs largely due to crazing. A ductilematerial fails largely by shear yielding.
For elastomeric applications as in TPVs, a plastic that can deform at arelatively low stress is more desirable than other commercially availableplastic materials; iPP meets this criterion (224). Moreover, for small strainsand short deformation time, iPP exhibits Hookean elasticity (storage ofenergy without any dissipation of the input energy as heat) and linearviscoelastic behavior (complete recovery of deformation on release of thedeforming force, although part of the input energy is lost as heat) for smallstrains beyond the Hookean limit before undergoing yielding at about 8%strain (224,225).
Figure 8 Chain folding in PP lamellae. (From Ref. 225.)
The most commonly occurring crystalline form of iPP is the a (monoclinic)form, which displays spherulitic morphology, as does the h (hexagonal)crystal modification but not the g (orthorhombic) or smectic forms. Thesespherulites are composed of radial lamellae in which the c (long) crystallineaxis is perpendicular to the lamellar growth direction (Fig. 8). iPP a
spherulites display a unique ‘‘cross-hatched’’ structure due to the presenceof tangential lamellae (Fig. 9), which is not observed for the lower density andlowermelting h and g forms. On heating or undermechanical stress, the h andg forms can be converted to the a form. A lower density, lower meltingmesomorphic crystal form that appears in iPP when it is fabricated underrapid cooling conditions also converts to the a form on heating (226).
iPP lamellae and spherulites are, of course, connected by tie moleculesthat are responsible for material continuity and therefore the mechanicalintegrity of the plastic. The initial Hookean response of PP observed at roomtemperature at low strain can be related to elastic deformation of the low Tg
amorphous tie molecules. The linear viscoelastic region that is observed forlow strain beyond the Hookean limit reflects enthalpic elasticity due to re-versible interlamellar shear or reversible intralamellar fine chain slip (Fig. 10)(224,225).
The fluctuation in amorphous layer thickness and the number andlength of the tie molecules throughout the sample can produce a high stressconcentration and cause material fracture at an average stress that is muchlower than expected for the tie chain density present (227). As PP crystalli-
Figure 9 Schematic of a spherulite with detail of ‘‘cross hatched’’ structure due to
radial and tangential lamellae. - - -,- represent ‘c’ axes of radial and tangentiallamellae respectively. The ‘c’ axes may be of the same size or be different in size indifferent directions. (From Ref. 235.)
zation conditions are adjusted to yield larger spherulite size, the materialbecomes more brittle due to the reduction in interspherulite tie molecules andthe fracture mode changes from intra- to interspherulitic rupture; a transitionfrom spherulitic yield to boundary yield is also observed (228–233).
In general, low-temperature, high strain rate, and plane strain condi-tions favor crazing of iPP.Under these conditions the various relaxation timesin the amorphous PP layers may be such that the stress cannot be transmitted
Figure 10 PP spherulite deformation mechanisms. Schematic diagram showing the
variety of deformation mechanisms operative in a semi-crystalline polymer, (a)interlamellar separation, (b) lamellar stack rotation, (c) interlamellar shear, (d)intracrystalline shear (‘fine chain slip’), (e) intracrystalline shear (‘coarse chain slip’)
(f) fibrillated shear. In bulk samples, these mechanisms coexist. (From Ref. 234.)
effectively to the spherulites, causing tie chain breakage and/or slippage thatresults in craze formation. Voids can also be created by lamellar fragmenta-tion. At high temperature and/or low strain rate, plastic deformation occursthrough unraveling of the folded lamellar PP chains. Radial lamellar orien-tation results in anisotropic spherulite deformation. Depending upon theorientation to the tensile direction, lamella can be separated, sheared, orcompressed. Voids can be formed within the spherulite through formation ofa fibrillar structure containing small pieces of lamellae due to coarse chain slip(Fig. 10) (225,228,229,234). A transition from ductile to brittle behavior of PPwas observed at room temperature as the strain rate was varied from 10�4 to90 s�1 (235).
The low a-crystal transition temperature (30–80jC) (218,236) of PPallows chain pullout from the crystalline segments, delaying the onset ofbrittle fracture. At very low strain rates, as in the stress cracking ofpolyethylene, the tie molecules holding the lamellae together untangle andpull out of the lamellae to cause interlamellar separation and brittle fracture(223). Although PP does not undergo stress cracking at ambient temperature,as PE does, when both materials are compared at the same temperaturedifference above Tg, their stress cracking behavior is comparable (237). Thelow a-crystal transition temperature observed for both PP and PE lowersthe energy barrier to stress cracking. This result also confirms the role of theamorphous component of the plastic in the stress cracking process.
Principle IV. Rubber Vulcanization
The fourthprinciple is that the rubber should be selectively cured to ahigh curestate for improved TPV elastic recovery, subsequent to the formation of anintimate rubber and plastic melt blend. The plastic should be minimally af-fected by the rubber curative. Some rubber and plastic grafting may providethe beneficial effects of improvedphase compatibility at the expense of reducedTPV melt processability. High molecular weight rubber with a narrow mole-cular weight distribution would allow a high rubber cure rate and cure state.
The curative that is added after formation of the desired rubber andplastic blend should not, or onlyminimally, affect the plastic phase and shouldrapidly diffuse into the rubber phase. In the initial stage of the rubber cross-linking process, a branched molecule is formed, altering the viscosity charac-teristics in the surrounding rubber domain. The branched molecule is also anucleus for the subsequent formation of a cross-linked rubber particle. Thebranched molecule can grow into a gelled rubber particle by entangling withand subsequently incorporating un-cross-linked rubber molecules into anetwork. It is desirable that several nuclei form simultaneously in the rubberphase so that the final rubber particle size is limited to the desired micrometerrange.
A rubber gel particle can have polymer chains extending out from itssurface, depending upon network microstructure. The gel particle can growby entanglement capture of un-cross-linked rubber molecules or by thecapture of gel particles that have polymer chains extending outward fromtheir surface. It is undesirable to allow the rubber particle to grow beyond the1–5 Am size required for good TPV physical properties in order to avoid thedecline in physical properties that occurs with a further increase in the size ofthe rubber particles. It is impossible to break up a rubber gel particle byintensive mixing of the polymer melt (238). Hence, before collision betweengel particles fosters uncontrolled particle growth, cross-linked rubber par-ticles that have only a limited number of polymer chains extending from theirsurface should be formed by rapid action of the cross-linking agent. There-fore, rubber curing should be fast enough to limit rubber particle growth,which is related to temperature and ‘‘mixing’’ speed. These parameters alsocontrol the diffusion of curative in the polymer melt. When appropriatelycross-linked rubber particles are produced, particle coalescence is avoided.Formation of small rubber particles increases the surface area and thereforethe surface energy of the rubber which is quenched by interaction of therubber surface with the linear molecules of the rubber-compatible plasticphase, at the expense of an increase in the viscous drag of the molten plasticphase over the rubber particles. Thus the cured rubber particles separate intothe plastic phase. TPVmelt rheological data suggest the presence of long-livedentanglements among almost touching elastomer particles, depending uponrubber content (239).
The amount of curative included into the rubber gel particles as it formsand the rate of curative diffusion into the cross-linking rubber network willdetermine the rate and state of cure. There should be sufficient cure sites on therubber molecule to form a network with ‘‘adequate’’ cross-link density for‘‘good’’ TPV properties such as compression set. A ‘‘lightly’’ cross-linkedrubber particle can still have ‘‘low’’ extractables (in the standard test used todetermine the amount of rubber molecules excluded from the rubber net-work) but may yield a TPV with poorer compression set than a TPVcontaining more densely cross-linked rubber particles.
Principle V. Morphology
Polypropylene/EPDMTPVs are produced bymelt blending of the rubber andplastic followed by selective curing of the rubber phase while the well-mixedblend continues to be intensively sheared. The fifth principle is that the mixingintensity during TPV preparation should cause fragmentation of the rubberphase into small (1–5 Am in diameter) cross-linked rubber particles and alsoallow plastic phase inversion. That is, the discrete or cocontinuous moltenplastic phase should become continuous as rapidly as the continuous rubber
phase is cured and broken up into cross-linked rubber particles. The processresults in 1–5 Am diameter rubber particles filling a molten PP matrix.Depending upon the rubber-to-plastic ratio and the melt viscosity (thereforemolecular weight) of the individual phases and the blending conditions, themolten blend morphology prior to cure could consist of 1) particulate rubberin a molten plastic matrix, 2) cocontinuous rubber and plastic phases, or 3)molten plastic particles in a continuous rubber phase (170,185,240). Afterdynamic vulcanization, it is desirable that irrespective of the initial blendmorphology, the final product should consist of fine, fully cross-linked rubberparticles in a plastic matrix. The variation in blend morphology of PP/EPRblends at 200jC and at a shear rate of 5.5 s�1 due to variations in the weightfraction and viscosity of the components has been reported (241). Owing topolymermelt shear thinning, the phase diagram varies with shear rate. Energytransfer between the phases duringmeltmixing is expected to bemaximized byphase viscosity matching. Phase melt viscosities should also have a significantinfluence on the rubber particle size produced during dynamic vulcanization.
Note, however, that material shear thinning narrows the considerableviscosity difference observed between the plastic and rubber melt viscosities atlow shear rates (239). Melt viscosity differences observed at low shear rateswould narrow considerably under the much higher shear rates for PPs withdifferent melt flow rates used in dynamic vulcanization. Also, thermal,thermo-oxidative, and mechanochemical degradation will lower materialmelt viscosity (177,242). Phase coalescence has been observed when themixing intensity of uncured PP/EPDM melt blends is reduced (170,191,243,244). Perhaps the factors just discussed are responsible for the formation ofrubber particles of similar size by the dynamic vulcanization of PP/EPDMblends with different degrees of mismatch between rubber and plastic meltviscosity at low shear rates (245).
It is reasonable to conclude that the mixing intensity should be at leastsufficient to maintain blend morphology and to facilitate curative diffusioninto the rubber. Also, the finer the blendmorphology, the easier it is to achievethe desirable TPV morphology. It is also evident that in the dynamicvulcanization of blends 2 and 3 (see first paragraph in this section) it is harderto achieve the desired TPV morphology (compared with dynamic vulcaniza-tion beginning with blend 1), with case 3 expected to yield TPVs with thelargest rubber particle size. In the case of blend 3 (continuous rubber phase),the curative should not cross-link the rubber almost instantly. If thishappened, the mixing process would have to grind the thermoset rubbergenerated into 1–2 Am particles, which is not possible (238). However, therubber curing should be fast enough to rapidly generate several rubber nucleiso as to limit rubber particle size, and the mixing should be fast enough totransport the gelled rubber particles into the plastic phase (phase inversion)before the undesirable agglomeration of rubber particles occurs. As the
rubber cross-links (the plastic phase should not be affected by the curative) itbecomes less compatible with the plastic, and hence it seems reasonable topropose that dangling chain ends and any loosely cross-linked network resideon the outside of the rubber particle, which should have a more tightly cross-linked core.
The most efficient packing for monodisperse hard spheres is a face-centered cubic arrangement with the spheres taking up 74 vol% of space, andtherefore a minimum of 26 vol % plastic for a continuous phase is required ifTPVs could be modeled in that way (246). Of course, a TPV would have adistribution of rubber particle sizes that are not spherical and are deformable.Hence, a continuous plastic phase can be established in the solid state in aTPV with rubber volume fraction greater than 74%. Nevertheless, very softTPVs (35 Shore A) appear to have a rubber continuous and discrete plasticphase morphology in the solid state (SantopreneRRubber 8211-35) (Fig. 11).Of course, it can be argued that ‘‘winding’’ continuous plastic ligaments canappear to be discrete in a two-dimensional scanning electron micrograph, butthe high rubber and oil content of this product would favor a discrete plasticphase over a continuous one. The rubber continuous phase in soft TPVs isprobably due to closely packed cross-linked rubber particles held together bythermoreversible entanglements and ‘‘spot-welded’’ together by the discreteplastic phase. It can readily be seen how increasing the oil and rubber levelswhile decreasing the plastic content of very soft TPV formulations canproduce a softer product with only limited processability and mechanicalintegrity. In the case of example 6 (Table 6), the higher level of plastic in theformulation permits continuous plastic phase morphology in both TPO andTPV.
The solid-state morphology of commercially available PP/EPDMTPVswith hardness greater than 35 Shore A is best described as consisting ofparticulate rubber in a continuous plastic matrix. There is an example of a 70Shore A TPV product (VegapreneR) that has a rubber continuous morphol-ogy (247), perhaps because it has a high rubber-to-plastic ratio and only aminimal amount of oil added for TPV processabilty and hardness control.The high percent rubber content (low oil and low filler content) may beresponsible for the touted improved elastic recovery of Vegaprene over otherTPVs of equivalent hardness (248).
Principle VI. Melt Viscosity Control
The molten rubber and plastic interfacial area increases tremendously onTPO-to-TPV transformation due to the presence of small cross-linked rubberparticles that fill the plastic matrix. The viscous drag of themolten plastic overthe rubber particles and rubber particle interactions would make a poly-
olefinic TPV melt almost unfabricatable without viscosity reduction by theprocess oil added.
The importance of the TPV interphase formation and adhesion betweenphases has already been clearly established. Even in the presence of paraffinicoil, which is necessary to control TPV melt viscosity, the trapped entangle-ments between the rubber particles and the PP amorphous phase are main-tained. The sixth principle is that a plasticizer that does not affect the adhesionbetween the rubber and plastic phases is necessary for the control of TPVmeltviscosity. More broadly, addition of plasticizer should reduce the TPVviscosity to allow easy processability without detriment to interfacial adhe-sion or physical properties. However, the addition of excessive amounts of oilto a PP/EPDMTPV formulations would eventually result in poormechanicalproperties. On cooling of a TPV melt, the oil rejected by the crystallizingplastic phase is absorbed by the rubber particles.
D. Rationalization of PP/EPDM TPV Elastic Recovery
Given that TPVs by definition almost always exhibit a continuous plasticphase, the mechanism of recovery has been an important subject of study anddebate. It is clear that in order for TPVs to show good elastic recovery, thefollowing conditions are necessary: First, within the two-phase structure theremust be an interphase that provides a high degree of adhesion between thephases. Second, the rubber phase must be sufficiently cured to behaveelastically. The rubber particle size must be small enough that it can bedeformed by the relatively thin polypropylene ligaments. And finally, theplastic crystal structure should be modified by the rubber particles to producean elastic plastic phase.
For excellent TPV elastic recovery, the iPP crystallites should beuniform in size and be uniformly distributed throughout the sample, asshould be the particulate rubber dispersed phase. There should also be anappropriate balance between tie molecules and crystallites in the TPV plasticphase. A large number of tie molecules will maximize the load-bearingcapacity of the plastic phase and allowmaximum extensibility before rupture,provided the tie molecule distribution and length are uniform, which wouldminimize local stress concentrations (249). However, the smaller the lamellaedue to loss of molecular mass to tie molecules, the more readily willpermanent deformation in the plastic phase occur, especially at elevatedtemperature or under creep conditions.
Thus, the nucleating effect of cross-linked rubber particles on the plasticphase (209–211,250) is beneficial for elastic recovery in PP/EPDM TPVs. Ininjection molded PP/EPDM TPV plaques, the plastic a-crystal nucleationdensity was too high for the observation of individual spherulites by polarized
light microscopy except at the specimen core where the spherulites weresparsely distributed. In the center of the specimen, where the melt cooling ratewas the lowest, a few h spherulites were observed (250). This is, of course,undesirable because themelting point of these crystals is lower than that of thea form and the h crystals are converted to the a form by heat or mechanicalstress. These factors would be detrimental to TPV elastic recovery. However,it should be mentioned that h spherulites deform in a more ductile mannerthan the a form due to the absence of the ‘‘cross-hatched’’ structure in theformer case (251). In fact, h-nucleated PP is marketed for high impactstrength applications (Borealis).
For a 73 Shore A PP/EPDM TPV, X-ray data suggest that not muchspherulitic structure is present and that the PP crystallites consist of separatedand/or fragmented lamellae (252). The percent crystallinity of the plasticphase in a PP/EPDM TPV is not much lower than that of injection moldediPP. The absence of a large number of spherulites in the 73 Shore A TPVsuggests a uniform crystallite distribution in the plastic phase. A considerablepart of the plastic phase network then consists of elastic (Tg 0jC) tiemolecules.
Our understanding of TPV elastic recovery has been advanced by thepioneering work of Inoue (252–255). When 73 Shore A hardness PP/EPDMTPV was stretched to 200% elongation in the path of an X-ray beam, notmuch change in the iPP crystallite orientation was observed, and theorientation was completely recovered when the deforming force was releasedimmediately after the 200% elongation.
On TPO-to-TPV transformation, particularly when the material rubbercontent is high, as already discussed, there is a considerable increase ininterfacial area between the rubber and plastic phases due to breakup ofthe TPO continuous rubber phase into discrete, micrometer-sized, cross-linked rubber particles. This causes the continuous plastic phase to form a‘‘cobweb’’ structure consisting of thin ligaments and thin coiled sheets, withgood adhesion to the particulate rubber phase dispersed therein. There maybe some thick ‘‘islands’’ of plastic in the structure, but when a tensile force isapplied to a TPV sample, it can be surmised that most of the deformation willoccur by ‘‘uncoiling’’ of the coiled thin ligaments and sheets, resulting in littleactual deformation in the plastic phase at the molecular level. The plasticligaments in the tensile direction are stretched, causing a compression of therubber particles in between ligaments and causing movement of rubberparticles in the stretch direction. The rubber particles should be small enough(f5 Am in diameter) to be readily deformed by the thin PP ligaments.Generally, the larger the rubber particles, the poorer the TPV elastic recovery.
Plastic ligaments that are perpendicular to the pull direction would be incompression. In continuous cyclic loading of soft PP/EPDM TPVs at room
temperature, where not much time is available for strain recovery, the TPVtakes a permanent set. However, after several cycles, the stress–strain‘‘hysteresis loop’’ shows little change whether the cycling is conducted aboveor below the elastomer yield stress (245). This suggests that some of the plasticligaments have shear yielded under the prevailing plane stress conditions (seeprevious discussion). Plastic and rubber adhesion is not lost during thisprocess, and, on release of the deforming force, the TPV recovers by rubberparticle springback (release of equatorial compression), the recoiling ofuncoiled PP ligaments, and the decompression of polar PP ligaments.
Huy et al. (256) studied the deformation of PP/EPDM TPVs by usingpolarized infrared spectroscopy. The plane of the polarized light was adjustedto be alternately parallel or perpendicular to the sample draw direction. Thechange in the intensity of the IR absorption bands peculiar to PP or EPDMrubber only, caused by molecular orientation, was studied. The study wasconducted on iPP, thermoset EPDM rubber, and iPP/EPDM TPVs undertensile deformation. The Herman ‘‘orientation function’’ increased consid-erably during the deformation of pure iPP to high strain, more so than thethermoset EPDM. However, the PP crystal phase orientation was less thanthat of the EPDM in the TPV. At a low deformation (50% elongation) in aone-cycle loading–unloading test, the orientation in both TPV phases wascompletely recoverable. In a TPV stress relaxation test to 1000 min at 200%constant elongation, the orientation of the PP phase increased whereas that ofthe EPDM rubber phase decreased linearly. When the modulus of the TPVrubber phase was increased by increasing the rubber cure state, the orienta-tion in the PP crystal phase increased as the PP ligaments probably experi-enced increased stresses while the harder rubber particles were ‘‘squeezed’’ontensile loading.
The work of Huy et al. (256) and that of others supports the emergingmechanistic picture of the ability of TPVs to recovery elastically due to therecoverable deformation of highly cured, highly dispersed rubber particlesand their influence on the crystalline structure of a continuous plastic phase ofappropriate ligament thickness. There is continued interest in elucidating themechanistic details of TPV elastic recovery (256–261).
E. Thermoplastic Vulcanizate Processability
Polypropylene/EPDM TPVs can be fabricated by any thermoplastic processsuch as extrusion, injection molding, and blow molding. The presence ofparaffinic oil in the compound can increase the volume of the continuousmolten PP phase by partitioning into the iPP melt from the rubber phaseunder the processing conditions, thereby allowing excellent melt flow. The oilrejected on crystallization of the molten PP is reabsorbed by the rubber phase.
TPVmelts are shear thinning, and a considerable lowering of viscosity can beachieved by increasing the melt shear rate over that achievable by increasingthe melt temperature alone. In extrusion, the die swell of TPVs is much lowerthan that of molten PP itself, with softer TPVs (higher rubber content)exhibiting less die swell than harder TPVs (higher iPP content) (262). Thisis due to the TPV melt plug flow that occurs due to wall slip, where the meltslips as it is pushed through the die, assisted by a thin film of molten PP in oilthat lubricates the die surface. The energy supplied to the melt leads to littlemelt deformation (and therefore low die swell) owing to the low frictionalforces between the melt and the die surface, and increased energy input to themelt simply results in more of the melt being pushed out of the die.Crystallization of the iPP on cooling of the extrudate ‘‘freezes’’ the smoothmorphological characteristics of the extrudate. TPV processability is en-hanced by the use of a plastic phase with a broad molecular weightdistribution.
F. Thermoplastic Vulcanizate Compounding
Thermoplastic vulcanizates, characterized by complete cross-linking of therubber phase in a continuous thermoplastic phase, more closely approach theperformance characteristics of thermoset rubber than any other thermoplas-tic elastomer. The improved oil resistance, compression set, and properties atelevated temperature qualify the TPEs for numerous uses. Important combi-nations include EPDM–polypropylene, natural rubber–polypropylene, butylrubber–polypropylene, and nitrile rubber–polypropylene.
Thermoplastic vulcanizates are precompounded by themanufacturer tomeet the requirements of the end user. The compounding step takes placesimultaneously with vulcanization and yields the required morphologydevelopment that serves as the basis for the thermoplastic elastomericbehavior of these materials as well as other material characteristics requiredfor specific applications such as fillers, including clay, talc, or carbon black;extending or processing oils; and various stabilizers. Other additives are usedto affectUV resistance, flame retardancy, or color. Rubber-rich TPVs are soldas intermediates or concentrates for additional formulating in a ‘‘letdown’’operation, which typically involves addition of thermoplastic, oil, and otheringredients to balance cost and performance.
Specialty TPVs incorporate plastic or rubber phase blends and addi-tives. The thermoplastic phase is modified to affect fabricability, appearance,and adhesion properties. Compatibilizing agents are added when needed(polar–nonpolar system) and are frequently based on functionalized(maleated) polyolefins or block copolymers. Block copolymers from polyole-fin-block-polyamide (263), polyolefin-block-polyurethane (264), and polyole-
fin-block-polystyrene (265) have been shown to be effective compatibilizersand adhesion promoters.
In addition, dynamically vulcanized TPVs have been prepared usingfunctionalized olefinic random copolymers as the elastomeric phase. The(grafted or copolymerized) functionalized olefin is cross-linked by addition ofmaleic anhydride during the mixing step, similarly to traditional TPVs. Otherinnovations include the use of combinations of a-olefins and cyclic olefincopolymers as the elastomeric phase (266).
VIII. TPE/TPE BLENDS: NEW AND OLD COMBINATIONS
Most thermoset rubber compounds are formulated, compounded, andfabricated by the rubber producer, but traditionally TPEs have been soldpreformulated by material suppliers to part fabricators and subsequently tothe end user. More recently, fabricators have been compounding in-house,especially for the SEBS compounds in the automotive industry (267) but alsoin the case of newly introduced TPVs (247).
Metallocene-catalyzed olefinic and styrenic plastomers and elastomershave entered the market that can be blended with other TPEs for enhancedperformance (267). In fact, one of the most significant trends in new TPEmaterials is the combination of different classes of TPEs with each other, forexample, blends of SEBSwith TPVs, to achieve performance and cost benefitsas well as a competitive position for thermoplastic compounders (267). Meltmixing of various materials, with or without added compatibilization,plasticization, or fillers, continues to contribute to dozens of new productintroductions every year. Elastoplastic compositions featuring a continuousthermoplastic phase with dispersed vulcanized EPDM as well as dispersedpolystyrene have been demonstrated (268). Formulated SEB(P)S/polyolefinblends from 0–60 Shore A with good processing and appearance character-istics have been introduced for overmolding onto polyolefins (21). Addition ofSBCs to TPVs increases the extensibility of foamed products (269). TPUsblended with SBCs, plastomers, EVA, and TPVs have created endlesscombinations with new characteristics. Indeed, the future of thermoplasticscompounders is rapidly becoming more complex than ever, with a myriad ofmaterial selections in TPEs, conventional elastomer modifiers, and new softerpolyolefins among the choices for engineering new materials. The future ofthese innovative thermoplastic elastomers remains promising; growth in TPEmaterial development and applications is expected to continue at double-digitrates for the next 10 years (9).
After over 50 years in the marketplace, many applications of TPEs canbe considered mature, others fully validated, and still others in growing and
emerging markets or applications. The breadth of those applications hasdriven innovation in markets across the globe and across industrial materials.Today TPEs are moving beyond rubber replacement, especially wherefeatures such as colorability and soft touch have found applications inautomotive and consumer products—anywhere the designer wants to addluxurious feel and color. Many current TPE applications such as consumerproduct grips would not have been possible with thermoset rubber or withmodern thermoplastics. Innovation continues in the twenty-first century, andmaterials scientists, industrial designers, and process engineers push the limitsof today to create tomorrow’s elastomeric solutions.
SUMMARY
The molecular characteristics of TPUs, SEBS compounds, and PP/EPDMTPVs have been described in detail, including the impact of the molecularfeatures on product properties and processability. The dynamic vulcanizationprocess has been described, and raw material selection criteria for theproduction of commercially viable TPVs have been established. The historyof dynamic vulcanization has been traced through the development of thistechnology. The improved elevated temperature recovery of PP/EPDMTPVsover TPUs and SEBS compounds has been rationalized.
Applications discussed for TPEs include the creation of new ‘‘rubber’’applications by redesign or replacement of existing materials including metal,wood, and soft plastics. Thermoset rubber replacement by TPVs has beenshown to be successful based on product fabricability (dimensional and‘‘flash’’ control), improved properties (stress relaxation, flex fatigue, density),price/performance balance, and fabrication economics. Thermoplastic elas-tomers can be readily fabricated by blow molding, and thermoset rubbercannot. Products that require a composite structure containing multiplelayers of material (plastic and TPE layers, for example) cannot be producedwith thermoset rubber. However, they can readily be fabricated by coex-trusion, coinjection molding, or multilayer blow molding of TPE materials.
ACKNOWLEDGMENTS
We are indebted to Dr. Garth Wilkes, Distinguished Professor of ChemicalEngineering, Virginia Polytechnic and State University, for his thoughtfulreview of the manuscript, which has enhanced its accuracy and clarity.Discussions with the following individuals are gratefully acknowledged: Dr.Geoffrey Holden, Holden Polymer Consulting; Dr. Jeffrey Koberstein,
Professor of Polymer Science, ColumbiaUniversity; Dr. EdKresge, PolymersConsultant; Dr. Yona Eckstein, Noveon; Dr. Gerald Robbins, Bay StatePolymers; Dr. Steve Manning, Bayer; and Drs. Dale Handlin and KathrynWright, Kraton Polymers. We also thank Marc Payne, Chief TechnologyOfficer, Advanced Elastomer Systems, L.P., for reviewing this work. CathyParker and Brian Gray cheerfully helped us with literature searches and inobtaining paper copies of the numerous references cited in this work. Wewould also like to acknowledge the contributions ofNormBarber andDr. LiliJohnson.We are also grateful to S. Rebecca Rose for manuscript preparationand for her patience with us during this process. Finally, we thank AdvancedElastomer Systems, L.P. for allowing us to publish this work.
REFERENCES
1. Gowariker VR, Viswanathan NV, Sreedhar J. Polymer Science. New Delhi,India: Wiley, 1987:2.
2. Encyclopaedia Britannica, Micropaedia. Vol. 10. Chicago, IL, 1988:223.3. Encyclopaedia Britannica, Macropaedia. Vol. 21. Chicago, IL, 1988:291.
4. Mark JE. Physical Properties of Polymers. Washington, DC: Am Chem Soc,1993:5–7.
tomers. New York: Marcel Dekker, 2001. Ch 17.8. Gilbert M. J Macromol Sci Rev Macromol Chem Phys C 1994; 34(1):77.9. Freedonia Group Study 1553, May 2002. www.Freedonia.com.
10. Holden G. In: Baranwal KC, Stephens HL, eds. Basic Elastomer Technology.Baltimore, MD: Am Chem Soc, United Book Press, 2001:546.
11. Carrabine L. Real World Insights from Consumer Goods Pros. Flowfront,
Wayland, MA: Moldflow, April 2003.12. Whitford M. Rubber Plast News, Oct 25, 1993.13. Thielen M, Schuller F, Balzer M. 3D blow molding today: an overview about
different systems which are established in the market today. SPE Proc ANTEC2001:903. Paper 470.
14. Koester L, Vroomen G. Coextrusion of EPDM and TPV hybrid profile.RAPRA TPE, Brussels, Belgium. 2001. Paper 13.
33. Petrovice ZS, Budinski-Simendic J. Rubber Chem Technol 1985; 58:685.34. Bonart R. Angew Makromol Chem 1977; 58/59:259.35. Leung LM, Koberstein JT. Macromolecules 1986; 19:706.
36. Peebles LH Jr. Macromolecules 1976; 9(1):58.37. Peebles LH Jr. Macromolecules 1974; 7(6):872.38. Seymor RW, Estes GM, Cooper SL. Macromolecules 1970; 3(5):579.
39. Lilaonitkul A, Cooper SL. In: Frisch KC, Reegen SL, eds. Advances in Ure-thane Science and Technology. Vol. 7. Wesport, CT: Technomic, 1979:163.
40. Clough SB, Schneider NS, King AO. J Macromol Sci Phys 1968; B2:641.41. Leung LM, Koberstein JT. J Polym Sci Polym Phys Ed 1985; 23:1883.
42. Abouzahr S, Wilkes GL. Ophir Z. Polym 1982; 23:1077.43. Van Bogart JWC, Gibson PE, Cooper SL. J Polym Sci Polym Phys Ed 1983;
21:65.
44. Ferguson J, Patsavoudis D. Eur Polym J 1972; 8:385.45. Zhrahala RJ, Critchfield FE, Gerkin RM, Hager SL. J Elast Plast 1980; 12:184.46. Chang AL, Thomas EL. In: Cooper SL, Estes GM, eds. Multiphase Polymers
ACS Adv Chem Ser 176, Washington, DC: Am Chem Soc, 1979:31.47. Wilkes GL, Samuels SL, Crystal R. J Macromol Sci Phys 1974; B10:(2):203.48. Wilkes GL. private communication.
49. Koberstein JT, Russell TP. Macromolecules 1986; 19:714.50. Hu W, Koberstein JT. J Polym Sci 1994; B32:437.51. Seymour RW, Cooper SL. Macromolecules 1973; 6(1):48.
52. Van Bogart JWC, Bluemke DA, Cooper SL. Polymer 1981; 22:1428.53. Sung CSP, Schneider NS. Macromolecules 1975; 8:68.54. Yang WP, Macosko CW, Wellinghoft ST. Polymer 1986; 27:1235.55. Wilkes GL, Bagrodia S, Humphries W, Wildnauer R. J Polym Sci Polym Lett
58. Ophir ZH, Wilkes GL. In: Cooper SL, Estes GM, eds. Multiphase PolymersACS Adv Chem Ser 176. Washington, DC: Am Chem Soc, 1979. Ch 3.
59. Kong ESW, Wilkes GL. J Polym Sci Polym Lett Ed 1980; 18:369.
60. Van Bogart JWC, Lilaonitkul A, Cooper SL. In: Cooper SL, Estes GM, eds.Multiphase Polymers. ACS Adv Chem Ser 176, Washington, DC: Am ChemSoc, 1979. Ch 1.
61. Heskett TR, Van Bogart JWC, Cooper SL. Polym Eng Sci 1980; 20(3):190.62. Martin DJ, Meijs GF, Gunatillake PA, McCarthy SJ, Renwick GM. J Appl
81. Nitzche SA, Wang YK, Hsu SL. Macromolecules 1992; 25:2397.82. Muramatsu S, Lando JB. Macromolecules 1998; 31:1866.83. Gabrielse W, Soliman M, Dijkstra K. Macromolecules 2001; 34:1685.
84. Gabrielse W, Guldener V, Schmalz H, Abetz V, Lange R. Macromolecules2002; 35:6946.
85. Sierra CA, Galan C, Fatou JG, Parelada MD, Barrio JA. Polymer 1997;38:4325.
86. Ohlsson B, Tornell B. Polym Eng Sci 1996; 36:1547.87. GergenWP,LutzRG,Davison S. In:HoldenG, LeggeNR,QuirkR, Schroeder
89. Barton AFM. Handbook of Polymer—Liquid Interaction Parameters and
Solubility Parameters. Baco Raton, FL: CRC, 1990.90. Ryan AJ, Stanford JL, Still RH. Polym Commun 1988; 29:196.91. Holden G, Bishop ET, Legge NR. J Polym Sci C 1969; 26:37.
111. Tucker PS, Barlow JW, Paul DR. Macromolecules 1988; 21:1678.
112. Aggarwal SL, Livigni RA. Polym Eng Sci 1977; 17:498.113. Mayes AM, Russell TP, Satija SK, Majkrzak CF. Macromolecules 1992;
25:6523.114. Winey KI, Thomas EL, Fetters LJ. Macromolecules 1991; 24:6182.
115. Hong KM, Noolandi J. Macromolecules 1983; 16:1083.116. Akiyama Y, Kishimoto Y, Mizushiro K. US Patent 4,772,657, Sept 20, 1988.117. Francis JV, Overbergh NMM, Lodewijk J, Vansant MFG. US Patent
5,710,206, Jan 20, 1998.118. Hall JE, Wang X, Takeichi H, Mashita N. US Patent 6,184,292, Feb 6, 2001.119. Shell Technical Bulletins: KratonRG7821X, G7702X, and G7722X.Multibase
and GLS Corporation websites.120. Southwick JG, Vonk W. Styrenic block copolymers. Chem Eng, Oct 1, 2002;
136. Hartman PF, Eddy CL, Koo GP. Rubber World, Oct. 1970; 59.137. Hartman PF. US Patent 3,909,463, Sept 30, 1975.138. Coran AY, Das B, Patel RP. US Patent 4,104,210, Aug 1, 1978.
139. Coran AY, Das B, Patel RP. US Patent 4,130,535, Dec 19, 1978.140. Coran AY, Patel RP. US Patent 4,130,534, Dec 19, 1978.141. Gessler AM, Kresge EN. US Patent 4,132,698, Jan 2, 1979.
142. Abdou-Sabet S, Datta S. In: Paul DR, Bucknall CB, eds. Polymer Blends. Vol.2. New York: Wiley, 2000:524.
143. Abdou-Sabet S, Fath MA. US Patent 4,311,628, Jan 19, 1982.
144. Coran AY, Patel R. US Patent 4,271,049, June 2, 1981.145. Sperling LH. Introduction to Physical Polymer Science. New York: Wiley,
2000:545.
146. Sperling LH. Introduction to Physical Polymer Science. New York: Wiley,2001:509–516.
167. Mader D, Thomann Y, Suhn J, Mulhaupt R. J Appl Polym Sci 1999; 74:838.168. Naiki M, Matsumura T, Matsuda M. J Appl Polym Sci 2002; 83:46.169. Jang BZ, Uhlmann DR, Vander Sande JB. Polym Eng Sci 1985; 25:643.
170. Pukanszky B, Fortelny I, Kovar J, Tudos F. Plastics, Rubber Compos ProcessAppl 1991; 15(1):31.
171. Dompas D, Groeninckx G. Polymer 1994; 35:4743.
172. Mehrabzadeh M, Nia KH. J Appl Polym Sci 1999; 72:1257.173. van der Wal A, Verheul AJJ, Gaymans RJ. Polymer 1999; 40:6057.174. van der Wal A, Gaymans RJ. Polymer 1999; 40:6067.
175. Karger-Kocsis J, Kallo A, Szafner A, Bodor G, Senyei Z. Polymer 1979; 20:37.176. Jang BZ, Uhlmann DR, Vander Sande JB. J Appl Polym Sci 1985; 30:2485.177. Dao KC. Polymer 1984; 25:1527.
178. Chou CJ, Vijayan K, Kirby D, Hiltner A, Baer E. J Mater Sci 1988; 23:2521.179. Lu J, Wei GX, Sue HJ, Chu J. J Appl Polym Sci 2000; 76:311.180. Bedia EL, Astrini N, Sudarisman A, Sumera F, Kashiro Y. J Appl Polym Sci
2000; 78:1200.
181. Grellmann W, Seidler S, Jung K, Kotter I. J Appl Polym Sci 2001; 79:2317.182. Wu S. J Appl Polym Sci 1988; 35:549.183. Wu S. ACS Polym Mater Sci Eng Fall, 1990:220.
184. Wu S. Polym Eng Sci 1990; 30(13):753.185. Pukanszky B, Tudos F, Kallo A, Bodor G. Polymer 1989; 30:1399.186. Da Silva ALN, Rocha MCG, Lopes L, Chagas BS, Coutinho FMB. J Appl
Polym Sci 2001; 81:3530.187. Kim BC, Kwang SS, Lim KY, Yoon KJ. J Appl Polym Sci 2000; 78:1267.188. Macauley NJ, Harkin-Jones EMA, Murphy WR. Polym Eng Sci 1998; 38:516.
189. Sherman LM. Plastics Tech 2002; 48(7):44.190. Barlow JW, Paul DR. Polym Eng Sci 1987; 24(8):241.191. Karger-Kocis J, Csikai I. Polym Eng Sci 1987; 27(4):241.192. Xiao HW, Huang SQ, Jiang T, Cheng SY. J Appl Polym Sci 2002; 83:315.
193. Mader D, Bruch M, Maier RD, Stricker F, Mulhaupt R. Macromolecules1999; 32:1252.
194. Dwyer SM,Boutni OM, ShuC. In:Moore EP Jr, ed. PolypropyleneHandbook.
Cincinnati, OH: Hanser/Gardner, 1996:220.195. Zhang L, Huang R, Wang G, Li L, Ni N, Zhang X. J Appl Polym Sci 2002;
86:2085.
196. van der Wal A, Nijhof R, Gaymans RJ. Polymer 1999; 40:6031.197. van der Wal A, Gaymans RJ. Polymer 1999; 40:6045.198. Martuscelli E. In: Karger-Kocsis J, ed. Polypropylene. Vol. 2. Copolymers and
Blends. London, UK: Chapman and Hall. Ch 4.
199. Rios-Guerrero L, Keskkula H, Paul DR. Polymer 2000; 41:5415.200. Zhang H, Wang J, Li J, Cao S, Shen A. J Appl Polym Sci 2001; 79:1345.201. Zhang H, Wang J, Cao S, Wang Y. J Appl Polym Sci 2001; 79:1351.
202. Lohse DJ, Wissler GE. J Mater Sci 1991; 26:743.203. Lohse DJ. Polym Eng Sci 1986; 26:1500.204. Chen CY, Yunus WMd.ZW, Chiu HW, Kyu T. Polymer 1997; 38:4433.
205. Ellul MD. Rubber Chem Technol 1998; 71:244.206. Ferry JD. Viscoelastic Properties of Polymers. New York: Wiley, 1980:374.207. Eckstein A, Suhn J, Friedrich C, Maier RD, Sassmannshausen, Bochmann M,
Mulhaupt R. Macromolecules 1998; 31:1335.208. Coran AY, Patel R. Rubber Chem Technol 1983; 56:1045.209. Inoue T, Suzuki T. J Appl Polym Sci 1995; 56:1113.
210. Inoue T. J Appl Polym Sci 1994; 54:723.211. Ishikawa M, Sugimoto M, Inoue T. J Appl Polym Sci 1996; 62:1495.212. Krulis Z, Fortelney I, Kovar J. Collect Czech Chem Commun 1993; 58:2642.
213. Argon AS, Bartczak Z, Cohen RE, Muratoglu OK. In: Pearson RA, Sue HJ,Yee AF, eds. Toughening of Plastics: Advances in Modeling and Experiments.ACS Symp Ser 759. Washington, DC: Am Chem Soc, 2000. Ch 7.
214. Wilbrink MWL, Argon AS, Cohen RE, Weinberg M. Polymer 2001; 42:10155.215. Thio YS, Argon AS, Cohen RE, Weinberg M. Polymer 2002; 43:3661.216. Xuanzheng W, Xiaoguang Z, Zongneng Q. 8th Int Conf Deformation, Yield,
and Fractures of Polymers, April 8–11, 1991. Papers 78 and 79.
217. GaymansRJ, van derWalA. In: PearsonRA, SueHJ,YeeAF, eds. Tougheningof Plastics: Advances in Modeling and Experiments. ACS Symp Ser 759.Washington, DC: Am Chem Soc, 2000. Ch 8.
225. Coulon G, Castelein G, G’Sell C. Polymer 1998; 40:95.226. Phillips RA, Wolkowicz MD. In. Moore EP Jr, ed. Polypropylene Handbook.
Cincinnati, OH: Hanser/Gardner, 1996. Ch. 3.
227. Peterlin A. J Polym Sci A-2 1969; 7:1151.228. Way JL, Atkinson JR, Nutting J. J Mater Sci 1974; 9:293.229. Freidrich K. Progr Colloid Polym Sci 1978; 64:103.
230. Ishikawa M, Ushui K, Kondo Y, Hatada K, Gima S. Polymer 1996; 37:5375.231. van der Wal A, Mulder JJ, Thijs HA, Gaymans RJ. Polymer 1998; 39:5467.232. van der Wal A, Mulder J, Gaymans RJ. Polymer 1998; 39:5477.233. Sugimoto M, Ishikawa M, Hatada K. Polymer 1995; 36:3675.
234. Zhang XC, Butler MF, Cameron RE. Polymer 2000; 41:3797.235. Dijkstra PTS, van Dijk DJ, Huetink J. Polym Eng Sci 2002; 42:152.236. Hu WG, Schmidt-Rohr K. Acta Polym 1999; 50:271.
237. Lustiger A. private communication.238. Coran AY, Patel RP. In: Holden G, Legge NR, Quirk, R, Schroeder HE, eds.
239. Krulis Z, Fortelny I. Eur Polym J 1997; 33:513.240. Manchado MAL, Biagiotti J, Kenny JM. J Appl Polym Sci 2001; 81:1.241. Romanini D, Garagnani E, Marchetti E. Paper presented at Int Symp New
246. Leutwyler K. Sci Am, Sep 14, 1998.247. Cretin JL. (Hutchinson Worldwide) Presentation at Automotive Elastomers
Conf, Detroit, MI, June 18–19, 2002.
248. Garois N. US Patent 6,028,142, Feb 22, 2000.249. LustigerA,MarzinskyCN,Mueller RR. J PolymSci PolymPhys 1998; 36:2047.250. Cakmak M, Cronin SW. Rubber Chem Technol 2000; 73:753.251. Aboulfaraj M, G’Sell C, Ulrich B, Dahoun A. Polymer 1995; 36:731.
252. Yang Y, Chiba T, Saito H, Inoue T. Polymer 1998; 39:3365.253. Kikuchi Y, Fujui T, Okada T, Inoue T. Polym Eng Sci 1991; 31:1029.254. Kikuchi Y, Fujui T, Okada T, Inoue T. J Appl Polym Sci 1992; 50:261.
255. Aoyama T, Carlos AJ, Saito H, Inoue T, Niitsu Y. Polymer 1999; 40:3657.256. Huy TA, Luepke T, Radusch JJ. J Appl Polym Sci 2001; 80:148.257. Okamoto M, Shiomi K. Polymer 1994; 35:4618.
258. Oderkerk J, Groeninckx G, Soliman M. Macromolecules 2002; 35:3946.259. Oderkerk J, de Schaetzen G, Goodeis B, Hellemans L, Groeninckx G. Macro-
molecules 2002; 35:6623.
260. Boyce MC, Socrate S, Kear K, Yeh O, Shaw K. J Mech Phys Solids 2001;49:1323.
261. Boyce MC, Yeh O, Socrate S, Kear K, Shaw K. J Mech Phys Solids 2001;49:1343.
262. Abdou-Sabet S, Datta S. In: Pual DR, Bucknall CB, eds. Polymer Blends. Vol.2. Performace. New York: Wiley, 2000:550.
263. Ouhadi T. US Patent 06414081B1, July 2, 2002. Also EP 0921153A and related.
264. Horrion J, Ouhadi T. US Patent 5,852,118, Dec 22, 1998.265. Dharmarajan NR. Eur Patent 0474927A1, published applicationMar 18, 1992.266. Abdou-Sabet S, Harrington BA, Brant P, Dias AJ, Datta S. US Patent
6,121,383, Sept 19, 2000.267. Eller R. Proc RAPRA TPE2000, Olefinic TPEs, SEBS, Elastomers, and PVC:
Intermaterials Competition, Markets, and Industry Structure Change, Paper 1.268. Finerman TM, Ellul MD, Abdou-Sabet S. US Patent 6,399,710B1, and related
patents, June 4, 2002.269. Wang DST. US Patent 5,939,464, Aug 17, 1999.
