MORPHOLOGY OF AMORPHOUS POLYMERS ANDEFFECTS OF THERMAL AND MECHANICAL

TREATMENTS ON THE MORPHOLOGY

0. S. Y. YEH

Departments of Chemical Engineering, Materials and MetallurgicalEngineering, and The Macromolecular Research Centre

The University of Michigan, Ann Arbor, Michigan 48104

ABSTRACTThis report covers the experimental evidence of order and structure in threepolymers, namely non-crystallizable atactic polystyrene (including amorphousisotactic polystyrene), crystallizable polyethylene terephthalate in its glassyamorphous state, and natural rubber in its near molten state. All the availableevidence from diffraction and microscopy studies strongly indicate that there is aliquid-crystal type chain packing order contained within a nodular structure oflimited size (— bOA) present in all three polymers. Effects of annealing andorientation give further support to the structure evidenced in the original poly-mer. The kinds of models which have been proposed for the amorphous state ofpolymers have also been discussed in order to clarify further the meaning of order

in amorphous polymers and the presence of truly disordered regions.

INTRODUCTION

Considerable study of the morphology of crystalline polymers in the pastdecade or so has led to the development of the chain-folded lamella as a basicstructural unit for both the undeformed' and deformed2 polymers. Althoughthere are still numerous questions concerning, e.g., the presence or absence of atruly amorphous phase inbetween the lamellae,3 the regularity of the foldstructure,4' and the origin and extent of interlamellar links,8 the recognition ofthis basic, lamellar structural unit in crystalline polymers has led towards amuch better understanding of the physical and mechanical properties of crystal-line polymers.

In contrast, little is known concerning the morphology of amorphouspolymers or polymers in their amorphous states (glassy, melt or solution).Clearly a better understanding of their morphology should also lead to a betterdescription of the stress—strain behaviour of polymers in their rubbery state; aclearer understanding of ductility, crazing, and yielding behaviour of polymersin their glassy state; as well as possibly answering some of the questions raisedearlier concerning the morphology of crystalline polymers.

A much more extensive review on the morphology of amorphous polymershas been written and will be published shortly.9 In this report we wish to reviewexperimental evidence of order and structure in three polymers, namely noncry-stallizable atactic polystyrene (APS) (including some results on crystallizable

65

P.A.C—31---I/2-D

G. S. Y. YEH

isotactic polystyrene (IPS)). crystallizable polyethylene terephthalate in itsglassy amorphous state and natural rubber in its near-molten rubbery state. Theeffects of thermal and mechanical treatment on the original amorphous mor-phology will also be presented to the extent that these effects are directly relatedto the original structure in the amorphous state of the three polymers. A briefreview of the various models suggested so far for the amorphous state ofpolymers will also be included, hopefully to serve as a base for furtherexperimentation and clarification of this controversial subject.

RANDOM COIL MODELUntil recently, the random-coil model has been generally accepted by most

polymer scientists as a fairly accurate description of the conformation ofmacromolecules in their amorphous state. This is schematically represented inFigure 1. In addition to the assumed total absence of order of any kind

(including I- or 2-dimensional), the single-phase model assumes numerouschain entanglements and a great deal of free volume. Being statistical incharacter, the model has been very useful in the original derivation of themolecular theory of rubber elasticity. Although the limited applicability of thistheory is known and has been pointed out from time to time, the questioning ofthis theory has only been taken seriously in recent years. The more successfuland recent molecular theories by Guth'° and BIokland have been developedby taking into consideration the presence of structure in rubber networks. Othersuggestions which indicate the incorrectness of the random-coil model havecome from the effects of solvents on the depolarization of light scattered fromn-alkane solutions,'2 on the strain birefringence of swollen polymernetworks,'3 '4and on the comparison of densities between calculated (based onthe assumed random coil model) and experimental values of various polymersin their amorphous state.'5

66

Figure 1. Schematic drawing of the random-coil model35.

MORPHOLOGY OF AMORPHOUS POLYMERS

PRELIMINARY REMARKS ON TECHNIQUES

Evidence of order in amorphous polymers can be derived from analysis ofx-ray or electron diffraction patterns either through the well-known Fourieranalysis of the experimental scattering curves or through comparison betweenthe experimental intensity function and a theoretically derived intensity func-tion based on some assumed model. Indication of possible order may also bederived from semi-quantitative analysis of oriented (and/or thermally treated)and unoriented diffractions of the same amorphous polymer, this being the leasttime-consuming of the three methods of analysis.

A more direct technique capable of detecting order and/or structure in theso-called "amorphous" materials is done in an electron microscope using eitherdark field diffraction contrast or phase contrast microscopy. The Fourieranalysis is then carried out automatically by the lens system, provided that theinstrument has the resolution capable of detecting the size of the ordered regionor the lattice spacing within a given ordered region. Neither the dark field northe phase contrast technique has been used to any great extent to derivestructural information on amorphous polymers, although the use of suchtechniques in studies on carbons has provided very informative data, e.g. ref.

To date most of the electron microscope studies on amorphous polymershave been on shadowed or stained samples. With virgin specimens the contrastis mainly based on thickness contrast principle.

The evidence of order and/or structure in the three polymers which we havechosen for this lecture comes chiefly from our own studies, based primarily onelectron microscopy and electron diffraction, and some on x-ray diffractiontechniques. Whenever appropriate, comparisons with available data by otherson the same or similar polymers will also be made.

POLYSTYRENE

Amorphous polystyrenes (atactie or isotactie) exhibit at least four diffuserings in x-ray or electron diffraction patterns. An example is shown in Figure 2,which is taken of an unoriented pattern. The corresponding Bragg spacings (9,4.78, 2.23 and i.26A) of these four rings are also indicated in the figure. The4.78A ring is the most intense of the four. The next most intense is the 9A ring,followed by the 2.23 and I.26A rings. The 2.23 and l.26A rings can be easilyassumed to result from intramolecular scattering. The distances between the ,-

C3 and C1-C2 within the molecule are about 2.23 and 1.26A. The origins of the9.0 and the 4.78A peaks have been elucidated by Krimm's study of orientedpolystyrene.'7 In the oriented pattern, of which an example is given in Figure 3,Krimm showed that the 9.OA ring and the 4.78A ring split into ares and orientalong the equator and the meridian respectively with essentially no changes ind-spaeings. The split in the 4.78A ring is barely detectable in Figure 3, as inmost eases. Krimm concluded that the 9.OA ring is intermolecular in origin,arising from scattering of atoms in neighbouring chains and that the 4. 78A ringis at least due to two different types of interatomic seatterings: those betweenatoms in alternate phenyl groups in the same chain and those between atoms inphenyl groups and main chain atoms in neighbouring chains.

67

H3Y .Y , .0 usdnio1H d b3niEido nd vI3fI nii1oq b nu noiiudhjgib IE1bE5I nia .ibui n-x mofi bru3 niiIj)I d b1u3 81n,nnoT bnl$ fIiI11E1

briE viin noiiudiflib IEin3rrri1qx i1i rr3wi3d bJ3rn iw normqmoo on riiknq nujrI lo bbom moil biBIu1E3 viu noiiudiiiib !Eibjn !E3ii10fli iw 2noi2IJknloD iin1bb on nii1oq 3nu!1Ei1 lo viu '1n-x mw1 io

01 !vtorn nuJ!oq lo ibio niA3Eq 1dioq rli oi i3q1 rliiw brI31 rIi1o ariin3ini di iEdi flio inioq bib .\i \ uErIrno18 ivwoH biIo r1i

iiJA .niJIoq DiJioi IJodq1OmE 1I3111o nl3flsq noiJ)ElThb noiib Lsziq'i A . •rniiEq uodqomz iJimi worl noi'iloq

Aoi modE fl idio r1J bru A iuodi j1 no hmrirnuiqx owi iii1j iud nui3un c1iEri bbEq 'dmobnEl E moi1 biqx3 fIEfli ir1irI dDJJfTI

13 vEd ol b3muE 13w irtrrig niEcb gr1huodrIi3n bi hE bkJo3 \rfJ ThIuiorrn3jni 3E13vE rIi Juods 1 AOI .3oru3Jib bbnix rno ivo niAmq

-udiiiib lEibEl rfi ni t13q gno1i 'i3v rli io1iclT) .n31'm(oq ni nEiib 13

riliw nii uTbb bo floiiE3o! 31i1 f1iB!qx OglE Aol JuodE JJ3 v11J3 floii .(n1flEq noiirniTtib 3113 01 AQ iuodE bo gniBq- nibnoqio

8

LC2TIJt 2 JJIIJJ pC 2flWW&I.!SCq pCJOtt'tpc qijjuiciiou b&gcuv 2RCJJ 2tnq JJ&AC pCCU cuicq Ofif LGCCIJIJA suq tpcCLO22JJJJJ(JU MJflJ jyrat CJCCtLOU25J tJJC 2tflL&UCC CSU pc qiLCCtjA wou!toLcq !UCJCCIL0U UJ!CLOL9bJJ2' t 9J20 2fIC2t2 t}Jt 'L 2HCJJ OLCL !2 qi2tfILpC ct nbouOIJC EIJJOtJJCV 21CP oLqcLcq LC!OU2 2pofIq LCACSj flJCW2CJ/iC2 ! qw cjqoLqw9tA (spicpc) boJAtALcuc ¶JLC bscgcq MItp 2OUJC OLCL UCSLIA btnisJICI to

JJJC &OAC SJJ&JX212 !IJqicStC2 tpst flJC IJJOjCCfIJC2 OL 2CtIJCIJt2 °L UJOjCCfJJC2 !UqJn2C LiU C&JJ pc qnc to !UICLUJOICCfIISL 2CSttCL!IJ 92 MCJJUCIJJpOnLIIJ CIJS!U 2CUJCUt2 9J20 COJJJC2 to SPOfIt 2f1C2t!1J tINt tJJC 4 \3yUJOJCCfIJSL UJoqCJ atnqC2 tJJt tIJC qiatmJcC PCtMCCU StOUJ2 !U bpcuAj toctha onLI2C tO S cprnn-tO-Cp&PJ cfl2tSUCC O9POfIt to ioy J1LtCUJJ0LC OUC uq2 LOWOL!w JJJC bSLSJIC, CpS!u b&quu iuqic&tcq p? yjit&qs atfiqica CSU 9J20 !ACq!a" LflJ !U IJJC qqpctiou bsttcLu !2 st iCt bw.trnijX !utLL!olCcnISL !UMCLC tg4a c{ suq ¶fl3 LC21JCCt!ACIA JJJ!2 9J20 iUqcStC2 tiNt tjJC 48woqcja tpc cj-çP BL-BL &uq j- qJ2tSnCc2 PCtMCCU tpc J2t &uq 3L bpcuAj L!u2UC!JJpof1L 2tXLCUC 2CUJCUt2 On UC!J011LIU CjJSIU 2CUJCUtr 411 tjJC &22fTUJC2CWCUt2 SIOIJ tpC CJJ&LU 92 MCII 92 9 LCflISI. &LL&UCWCUt 0 tIJC UC&LC2tCfILAC2 cacnpitcq LLOUJ LUOCj2 M!C S22fUUC 911 OLqCLA bscjcru oJtpc 2tALCUC

Tx-L&A qqjircuou bsffcLu OOLICIJfCq swothpon2 bojA2fALcuc 3fLCfC qrLccf iou ACLfJC&J

2jJOMC 141St tIJC CXbCL!LJJCUtSJ !UtCU2!tA CflLAC2 9LCCq ACLA MCII M!tP !UtCU2!tX

atuqica °L -CV bBLe SIJq b-4-qCL1AstAC2 oj 9t9CtC bojA2tALCut50 jctqJ31OL11p9f1 t V1\ M92 JStCL ACLICq pX y9t9q92 qctsqq C!CCtLOU qftL9Ct0u

IPC 2flCt!OU O 911 0LqC14? bscriu oj boJX2tXLCuC CIJS!U 2C&IJJCUt? pA

IAIOB'LHOFOOA Oh VVIOB'bHOfl2 bOFAYIEK2

G. S. Y. YEH

Ordered regions as revealed by electron diffraction22First I would like to describe some changes in the diffraction patterns

observed when freshly prepared thin films of atactic (three molecular weights:4800. 5 1 000 and 1 800000) or amorphous isotactic polystyrenes were irra-diated with increasing amounts of 80 kV electrons. Figure 4 gives an example

Figure 4. Densitometer tracings showing the changes in the 4.78 and 223A maxima as afunction of increasing crosslinking from bottom to top22.

of such changes in densitometer tracings. The major changes occur in the4.7 8A ring; changes in the 9A ring were not monitored because of its closenessto the main beam and because of its much weaker intensity to begin with. Notonly does the location of the 4.78A ring change, corresponding to an apparentincrease in d-spacing from 4.78 to 6.38A with increasing time of exposure tothe electrons (increasing degree of crosslinking), but there is also evidence ofincreasing line broadening as well as a decrease in intensity (after correction forthe background scattering). These changes appeared permanent upon reexa-mination. No detectable changes, either in intensity, d-spacing, or line profilewere observed in either the 2.23A or l.26A rings.

The observed effects of electron irradiation clearly indicate that the 4.78Aring arises from some kind of liquid-crystalline chain packing existing inamorphous polystyrenes. The ordered chain packing can be disturbed uponcrosslinking by electron irradiation. The increase in disordering of the paracry-stalline lattice results in the expected broadening of the peak, the expected shiftof the peak position towards the main beam, as well as the expected decrease inpeak intensity. Such changes are not expected to occur if the diffuse ring is due

70

20

a51ELMYJO UOHSIOMA 1O YOOJOH$LOM

0111 od ot bnudl asw boobni as nhoJis3a 01311-rEq bigi-i d boenso oDnoDhoini ol .rrigno ni TsIuooIornwun; ots dorfiw agnhA ô. 1 orb bns A ES orb rbiw oaso

DiJOSIS orb lo irigiow BIIJOOIOU1 orb lo aao1btso mlirnia otow agnibrrft oriT -ia'1oq oiiofloai orb rbiw borlisido 010W allrraot orrrsa odT .boau orroviailoq

oliosis 008 4 WM 10 oaso orb cii boson asw noilqooxo ciA .0000 WM orro1 risrb toq-rsrla lsdwornoa od 01 botsoqqs Asoq A8V .ts orb isdi cii ono11(1ai<1oq

orb luods boriisrnoi agriit Aô.i orb bns AE orb ;atorbo cii bovtoado oaodb .aonorsa1oq torbo cii brwol oaorb as orrisa

qoaaoioim f10t13313 bbft *isb 'd b3!s3v31 as anoisi bstnbio lo 3sia

010W lsrb aoibuia bloui *rsb noituloaot rigid ornoa odrroaob ol rlaiw I IxoM

3rI1yJaioq 3;bsJn lo mill nith bswobsrbmi us b dqsiotñm noU33is bloft tisU .t wntgR .9000008 1 .W.M)

'V

00 'S

200 'S

I

CLcnhIA pqocq fJJIU JuJ 2bCCIWGU2 /1GLC wqc LLOW 2CAGL&J JfCfiC

oqnai. (CZflLC p? paqoMu tGcpu!dnc

IUCLCJ2G ill 2IGqGUcA 0 fiG pic 0LGLG LG!0U2 fO JOUJGLffG LG2f1JffU HI fiG bbiLGufthGCILIJGU H0MGAGL fG2G JLGL LGI0U LG bL1WLIJA CIIGq pA fJJC fGLJ-LGi0U (—'20 fO Hi iG) CU JO pG GGU Hi fiG fluJOLbp0f1 !0fCfC1WC pOfff 3\ LOL Of IAbG2 o b0JA2fALGUG y&uA JfLGL OLqGLGqAGLarfJ FLow owc LfUqoUiJA GJGCfG qrucfiu LGi0U2 Hi fiG fL( cjqbLoqnccq pA paconuq qIjI2G CfffCLIU jjic 21G oj 0LfGLG LC!OIJqGuGq pupfJA qftLrcfiU 2bof2 &fiu2f LGJJfLAGJA ITUILOLW q&Lj cjqiii uq Q jjic 0LqGLGq LGJ0U LG Lccq!f A 02GLAJGb0JA2fALGUG2 U °L V'JM 1 800 000 uq U 1b2 O V!M 220 000 LG JO/1U

ExwbJG o qm jjq tWG2 opwiucq pow f/10 nupqocq uioLbEJo1f 2

boJX2ALcuc-c jcq GICCfLOU wcLouthp °L flJ flU2pIiqOMcq 1}JIU JJW 01 mJJ0LbJJ0fl !2O&C!C

2fnqiG2 CJLLIGq 01ff pA OfpGL2UCLGJ2UJ qi20LqGLHJ 0L fC 0LGLG LGIOIJ2) iuq LL.ow ru1A22 °L qqjl&cfiouGLJ1GL LLOUJ OlE 01111 JLHJ1A212 OLfPG CUG2 iii jL&Cf IOU b&ffGLu2 (cn2Gq pAGffGL fpJ Jy JJJG LG2IIff2 COUUJi fiG bLG2CUCG 0OLGLG LGIOU2 2flG2fGfWICLO2CObG GdfllbbGq MifJJ PGW fiJf qCAICC &uq fG2fG fO 1 LC2OJIIf IOU OL

btolccf fiG I1JJG JJiG CXbGLIWGUf2 MGLG CLLiGq 01ff 113 bPHI!b2 GICCfLOUC&LLIGq 01ff 013 fJJ fJJG bOJA2fALGUG2 pA 1f2iU f boLf IOU °L fiG 'V8 L!U fO

0 2 K AEH

2OO

5IHMYJO 2UOH5IOMA O VDOJOH5LOM

OiiOBiOai 3110 brin 000 008 1 bnz e000 1 a !008 Q 1 0O8 1 WM 'lo a3n3ta'(10q 3luixoi 'miubon izaiq'(i B boworia daidw lo Liz OO00aa WM lo onovJa'ioq

nornnqa boinoiio on ni oidnrlaiugniiaib oiom ri3um 3m a3lubori 3riT .( -si5k)

.bowobnda-tq 3noviayIoq 3!i3E1s 1o mill nirii bsriDJalJenU .' sitmR

auth awobnda sth bnidod nosa 3d oafs nns asiubon adi lo nnM .(? siR) lo noiinlunnmg orb ol sub sin aslubon bo'nsado orb izth iiiidiaaoq orb gnivomo-i

mnlirnia 'igniaiiqmua ai asiubon orb o oxia sriT .boau inimsinm gniwobnria sth

.b,wobnria-jq ,izi'M no WOOi bsriw5na 3nv2'ioq DiiDflB Th miff niriT 2 LV

bcfjcLu OpiLJCJ WW I1J JLC1!WflL O fJJC Ouc 4JOMU (Hi fiG L!4Jr;boJX21?Lcuc JfCL LJLJCfJI1J U I LOL OUC p0111 JJJG IUGLf J0M Hi GJGCILOU qijL1cfio1J

BLIIU JGJq CJGCILOU WiCLOU1bp OflJ flulpqOMGq 0Jflf1OU C&2( (P!IJ EJIW 01 !11C1!C

ywcic bo1AfALdu pocq icLA jirjc iipjc cpuc iuA i,pcu nJuGqcqcJJuc U IJJC uoqni CXtIILC qnLu JU1JCJJ!UJ

IP i cuA fIG woccnj GfMGCU fG uoqnjcMCLG J9L JOM boJA2fALGuG2 f 2IIJjG uoqnjc JJJ1A COU2!f O CAGLJJ wojGcnJc2UOq111L (oi. fGI4iJLA) 2fIJICff1LG MIf fiG WOJGCflJG2 Lf1UUHJ GfMCGIJ flJGWboJ?ALGu 1LG CxbGcfGq fO IJJAG WOGCfl1C2 C0U2ifiIJ O 2GACLJ 0 fJJG2GJJJGLGj9LG MG CJIJ coucjnqc jL0W pop 2CP0OU2 fnqiG2 uq 0f1L fpJf pip!M .iOOOO \f qifwGfcL o ic nJ GdfliMJGuf Aj oi. OUJA fOOOOboJAfALGuG 'i VM ooo iuq flJC 0fGL MJ2 rpon y u boj?2f?LGuG ip

AGLG MGLG LGboLfcq ouc i iponr ov l9 2fLflCWLG 2GGU iu

fGLfiJL? 2fLfICffLLG ILL 2WiUCq fJJIU UIIJJ2 0MJLi0fl2 W0JGCflJJL MGipf b0J?fALGtJGMOLfC 1MJI1P1G LOL cowbcLi20u cpooi LcboLfGq 0IJJG ML1Lf IOU LU LG OL pLLGLIG2 oj fJJG AM oj pc bojAfALGuc fl2cq OL Ip0f1f 30y lU fG OULA OfIJGL

2 K AELI

F-

MORPHOLOGY OF AMORPHOUS POLYMERS

below or above T9. This was expected since the polymer is noncrystallizable.Consequently isotactic polystyrene was used. Unlike polyethyleneterephthalate26 or polycarbonate27'28 which showed a tendency for the nodularstructures to aggregate or grow in size, isotactic polystyrene showed hardly anyvisible changes when annealed below Tg. On the other hand, distinct changeswere detected when IPS was annealed above T9. An example is given in Figure9. The various stages of transformation from a nodular to a spherulitic textureare clearly discernible in this micrograph. Other results showed that thenucleation of a spherulite appears to begin with the crystallization of a liquid-crystal like nodule, or a group of nodules merging together to form a few fibrils,which eventually fan out from the nucleus by additional incorporation ofmaturing nodules.

POLYETHYLENE TEREPHTHALATE (PETP)The major publications concerning the presence of order and/or structure in

the amorphous state of this polymer are by Yeh and Geil26 on the glassy state ofPETP and by Ermolina et al.29 on the molten state of PETP.

I wish to summarize their findings in the following.

Order as revealed by electron diffraction29

Ermolina et al. analysed the electron diffraction patterns of PETP that wereobtained at temperatures above and below the melting point of the polymer bymeans of Fourier analysis. They found that the experimental radial distributioncurves are similar, each containing several maxima. The one at 4.45A found inboth crystalline and melt specimens was assigned as intermolecular in origin,due to scattering between atoms in neighbouring molecules which they suggestedhave a parallel packing arrangement. The size of the ordered region was notmentioned, nor could it be determined from their data. The disordered regionwas not mentioned either; but presumably it belonged to the boundary regionsurrounding the clusters of parallel molecules.

Ordered region as revealed by electron microscopy26Yeh carried out experiments on unshadowed thin films of glassy PETP (MW

15000) at room temperature using bright field and dark field techniques. Anexample of a pair of bright field and dark field electron micrographs taken of thesame area is shown in Figure 10. Regions of fair contrast of about bOA whichindicate the presence of order and/or structure can be seen in both micrographs.The dark field micrograph was obtained by using a portion of the innermost,which is also the most intense, diffuse ring with a Bragg spacing of 4.5 A. Thediffracting regions are suggested to have a paracrystalline packing of alignedchain segments with a constant chain-to-chain distance of about 4.5A.

In a shadowed specimen a distinct nodular texture was detected (Figure 11),the average size of the nodules being about 75A. These structures are present inboth bulk and thin films, and have been confirmed in a separate study byKlement.3° The nodules have a fairly distinct boundary, indicating the majority

75

(W O 1WOLbpO112 bo?cfpA1cn fcLcbpfpc(cjipr jGJ (fob) &LJqL cq (poffouJ) GICCILOLI WICLOL1bJ O1RJ flLJ}J1OMC flJIU

U K AEH

fpL0fl&pon pc &uuc&jcq abcciwcua' LP! 2F1&G2f2 f&i 9 2RP2WUfJ&J &UJOFItJI oj(2oo 4D !000y !' 21SC) OJ, IJOqfljSl 2ILOCITILGa (jthcn y &uq fl) LC2JfGOCCI1LLG !11 fiG wothpojofl °L fiG 9JJUG9jG abcciiiiciia' J4UUJGLOIT2 &&LG&fG20LGL GfMGGU baucjjcj cp&iu 2GUJGU2 M!1HJ fiG tJoqfljG2 O11JGL C9IJGiuqic&fiu 9U !UCLG9aG !11 IJJIGLUJOJGCflf9L fP9I !UCLG92G JU fiG b9cjciuq'ro b&UGLU If GCtfIJJG apsLbcL MGU asujbjc suuc&jcq f QOC

9 2JJJJf 2JJ&LbGUiIJ Of f}JG fR y L!U congq pc qGfGCfCq Hi fJJG GJGCILOUbou &UUG&J!U 9f fG!JJbGLSffILG2 &Oflf 100C pcjoit \]( 0j bELL !2 &pOITf

UOqflJ&L 21LHCWL25US \ V 21i1CC LCbJ!C& ojjcujoLbpon2 njjc bojAcfpAJGuc rcLcbpcp&J&cc 2JJOM!IJ fJJG bLc2ducc 0

ponuq&LA' JUCG fiG JJJOJGGHJG2 &LG J0IJGL fJJ&U fiG C{J&WG(GL 01, fiG UOqfIJG2q&Ljc pGJq 2fflqJG2 ocqq f1&&G2f fiG bLcaGucG 01, 2OUJG 1,ojq-pscya &f fiG U01TJ9L10&GfJGL /tIfJ fJG b9uqjGj bacriu& 01, &jiuGq CJJ9JU 2GUJGUI2 92 !uqicafGq 1,LOUi01, CJJ&IU 2G&IJJGUf 2 pciu G22Guri&jjA CoUfaiuGq t/\!fP!U !AGU Uoqnjc JjJI2

V108'bHOFOOA Oh V110ffb110fl2 bOFAV1EI2

H

10 1}JC OU2GI flJ W021 b0JAWCL IIJ WGCPIIJJC5! LC!0W1IOU LGUJHIJ2 fO pc 2CCIJ

fcwbcuc1nLc 1MPC1P 1p!2 OLJGf 0LJOflJTL WOACIJJCLJI C31J pc Lcjtcq

\'5 2flLCC 1LI1C11ILG 01 w0Lbp0n bEib 1CL fLJIJG*JI1J TI Q0C l0 c

uq pA 2Jcurnuu niq Q'J58 0L jrA bojAcLpouuc nbou &uucjiu ufIOU2 ILG 10 2OIJJG GXIGUI 2IwijL 10 IIJO2G LcboLlcq pA ELLJ!( VICLLUJOJGCIIfm. W010U 12 0CC11LL!LJ icu u 1GWbCLfflLG2 pcJoM JJJG2G O2GLA

0 K KEH

soOL WGcji&JJiC&j 1LGUWGLJ1

uoqnjsL GXfflLG fG uoqnca patc c jidniq-jijcc cLAaralriuc OLGL bLroL o flJGLIJJ&JJAG coucjnqcq LOW pc GxbGuuJGU2 rpig jsaaA &uJoLbpona bELL P'2 '

fl 214CG LGbI!C1 °L W0Lbp0fl2 P'IK bEth &IJCL &U1JC&J!IJ U QOC LOL Q qAa1(

&11A bLG2GUf SIJGL f}JG iUifHIj 2ILGfCJJ3sLGACLj1IJ pscjc O f 3-qiwGualolJ&j awccç;c-çAbc JidnJq-cLAajsJpuc OLGL 04&!U-LG211G1CP1U flJG JJGf-Cf 2bCCiWCUa fiG 3-qimcua!ouq OLGL GCUJG qcarLoAcqCf &uq LG24TLGfC OU S arIpafLsjG (xJth%1.6 Q) if 12 (4 1LJfGLG2j O JJOjG jJJFJ( SJJGLbLioL 10 JJGSI 2Gf (xci.c SUGL pcar act sr 5@00C (\LThS )suq SUGL pcs

JJJG uoqnjs 2iLTJCUILG2 MGLG SSIU WTICJJ IJJOLG qiariuci iii OLJGUIGq 2LGC1IJJGJJ2

V1OffbHOFOQA OL V1AIOffbUOfl LOFAVIEff2

2OO

80

tuG qUrM qILccriou13'thLE; %• JiflU !lW 04 bELb qLcMu &4 (20C 1) t 1!1JJG J1 oLiwiJ JGLJrp 1JJG TLL0M UqIcI1C

0' K AEH

F—

L1'

gjOL!IUJ 2IiGCP qiLGciou

V cpcq P!U JIw TJL boqLMiu ioow ! qacpou biLJjJcj o pc

tI.ccpq fjJiU i' O bEib 1ISL pc ctu r 3Q00C OL j wwfTrc

OffbHOFOOA Oh VVIOffbHOfl2 bOF)J"1FI

Jooo

G. S. Y. YEH

NATURAL RUBBERNext to polyethylene, natural rubber is probably the second most inve-

stigated polymer. Yet very little is known concerning its morphology at roomtemperature in its most employed state.

Because of its high molecular weight (in the millions), it remains strongwithout crosslinking iii its near-molten state at room temperature (Tm of naturalrubber is about 25 to 28°C). Consequently structural studies can be conveni-ently carried out at room temperature in its amorphous molten state.

Structural order as revealed by diffraction techniques33Most of the x-ray diffraction studies on natural rubber have been on

stretched, crystallized samples. Simard and Warren33 are probably the onlyones to have analysed in detail the x ray diffraction of unstretched naturalrubber obtained at room temperature by the radial distribution method. Theyshowed that the corrected experimental intensity curve contained one veryintense peak at sin 0/).—0.104 and two other much less intense peaks (at sin0/). 0.22 and 0.41 respectively); the latter two are generally accounted forby intramolecular scattering of the carbon atoms. C1-C2 and C-C3. Theintensity of the innermost one was about 7 to 10 times the outer two. (Theorigin of this peak is still considered to be unknown; its intensity, however.generally decreases with increase in orientation and finally disappears altoge-ther upon further cooling in 'fully' crystallized stretched natural rubber.34 Themeasured crystallinity for a 'fully' crystallized natural rubber is, however,seldom more than 30 to 35 percent).

In the radial distribution curve obtained by Simard and Warren, at least fourpeaks at r - 1.52, 2.68, 4.0 and s.oA were observed. The first three wereattributed to the first, second and third neighbour carbon atom distances. Thelocation of the fourth peak was explained by its high intensity (concentration)due to an atom's nearest carbon neighbours in other chains, i.e., the distance froma carbon atom in one molecule to the first concentration of carbon atoms inneighbouring molecules which is about 5A. This high concentration of scatter-ing, as pointed out by them, is responsible for the very strong peak at sin 0/).0.104 in the experimental intensity curve. They did not explain, however, howthat high concentration could arise if neighbouring molecules were not packedwith any order. However, such concentration could arise if the neighbouringchain segments assumed a near-parallel alignment to one another.

Ordered regions as revealed by dark field microscopy3As we have indicated above, the most intense diffuse ring may be caused by

intermolecular scattering of atoms between neighbouring chains that are more orless parallel to one another. This was proved by high resolution dark field electronmicroscopy using a portion of the innermost ring with a Bragg spacing of 4.5A(Figure 17). A similar image was reportedly obtained by Ban3' in thin sections ofcrossl inked natural rubber. The interpretation ofthis dark field image is the sameas for the ones obtained in polystyrene and polyethylene terephathalate. The sizeof the ordered regions averaged about 30A, which represents a time and space

82

a5I3MYJO UOH51OMA 1O YDOJOH5IOM

1J31uiEno 3iEi iddui cIi th noit niiasThib iom 1dEdo1q £10 s1vI .icfdui

ni 3luix3i lJ3Iubon b !i!dEi 13r1 1oifnodB! mO 111 Ii3IJJ \Iifl3O1 ioM -ioq iddui Lfnuisn bwobBrk-jq bni3 bth -OO iDr1iI ebimob-bio

r1T .Iqrrrnx rno Q' bilE % .muiqmJ moom nm iovwoH .noim noii3smThb rli n&Ii mmJ31 1iDnijgib ei (AOO1 —) i1ubon

1ubon 3rli izrb bwor1 rbuJ noiizi1!Eim ni aiLum r1ir1w nirbimi noqu a3vf3m3f!i 1ubon r1i as mj ,rfi iuodB noi1 b3m3blo b3rIiEino

.(c3

.i,cJdui !i-iun b rnlil nidi bwobdnu ru lo r1qiigoim rroii& bbfl *IEQ . ¶r5t

mwot oi ni1ooD noqu iniiib iom rbum m&M 3liflx3i u1ubon r1T .1 'Vk ni nworI i !qrrux nA .1uiE1qmi

ni!om rfi iii tnmq i 3muiufl lElubon di iBdil noiiE3ibni m3clionA nd bud rl3icIw mill nith £ niillEim3 d &11J cJ b3vomq asw mddin icmuisri b!o rb n3miq bsi!1ui13nu !uniiio cIi nI .niIoo oi 11c\bu1o33b bIos .( \. V1 ) &ubon r1i sd b3iqu33o noi-i r!i biov oi b3bni 3IMimEq

inmq d ol bvido iw gnoim b3biov-blog is1imi fioiiEittEi2'13 floqU iu!1msI udli iu oi bni ir1T .( 3uIIomu1 lo rIinI r1i nofu 3iin1oi lo noiiui[LEi13 lo 3th ni as eilom i{i molt noiiusiff'çi rf molt 1unothutoq mo uiurir1qiii 3rr3brj1iioq t2Ioq

&iHw aIubon 3di 10 ciiA3uq bnu noiisxiI1ui13 d sIq aIui ui .u1bmuI nu1IEi213 bv13do 3rIi ni i1um

fcxfnLc: (0) CW4 OU M&fCL fIL10CG (P) U'Ô IP!0 JW 01 UOIhLOJ LnppcL f-}J9qOMGq Of WOW fCWbCLOIIILC 10 1O111 1G uoqn&L

— IOO\' 10\W IP!0 JW 01 UO111LO LflGL oq-qccoL&fcq Of LOOW IGWbGLO1nLG n-AOiqcq LGIOU

0 2 A AEH

(a) (p)

I Io-J '

82

& LOOW GWbCL&WLC it cLA1&IIis&tJou &uq O2O 21&iUcq ci\y&. CLA21&JJ!J11!O1J35 W&WL&l Lf1GL cLA2r&Jscq pcjom LOOW GIJJIJGL&UILC IJJC 1P! 111W /t&2 fl-qCCOL&fCq

f}JG uoqn&L LC!OIJ2 O2O at&iucq•%WL 3y flUCLA?f&JJisGq LGIOIJ OLU&tflL&J LRCL &UCL COOJHJ jt & I0MCL fcWbGL&HLG ajottni

4tWbCL&4TILG33W JLP!U 111W 01. IJ&fllL&1 LflGL uLGICjJGq fO &f)Oflf .OOo\° &uq o2O u&iucq &f LOOm

V'IOKbHOFOGA OIr VY1OKbHOfl bOFAY'IEff 2

_-r-T;—TI

o.J

G. S. Y. YEH

STRUCTURAL MODELS OF AMORPHOUS POLYMERSVarious structural models have been proposed ever since the first suggestion

of order in amorphous polymers by Katz in 1927. Some are based on structuralstudies while others are deduced indirectly. Some are also more specific thanothers.

Kargin proposed a bundle (or cluster) theory in 1957 based partly on x-raystudies of polymer melts showing high intensity of intermolecular scattering;and partly on bright field electron microscopy observations, showing spheru-litic fibrils which remain undestroyed after destruction of crystallinity as well asfibrils that formed upon mechanical orientation.35 Although the interpretationof such data, especially the electron microscope observations, may be some-what questionable, the idea of parallel alignment of long chain macromoleculesin the amorphous solid state appears reasonable and analogous to the swarmtheory proposed originally by Bose in l907, and promoted later byOrnstein.4° According to the swarm theory the interior of a nematic liquid iscomposed of clusters of molecules (10 to 106 molecules) with definiteboundaries.4'

In Kargin's bundle theory the parallel alignment of chain molecules is theessential feature; the size of the aligned region can fluctuate from a few segmentlengths to lengths greater than fully extended chain molecules. The order isstrictly intermolecular. The disorder is primarily in the boundaries between thebundles.

Hosemann42 suggested another model for the molten state of polymers(mainly polyethylene and polyethylene oxide). It is a one-phase paracrystallinemodel, based primarily on x-ray findings of disorder within lamellar crystals.With increasing temperature the disorder within the crystal is increased untilfinally at the melting point the lattice no longer has a 3-dimensional crystallinelattice; the chain segments, however, remain essentially parallel to one another.The folds originally present in the crystal are also suggested to remain in themelt. He pictured the melt as consisting of highly disordered crystals (seeFigure 16 in ref. 42).

Yeh35 proposed a two-phase folded-chain--fringed-micellar-grain model forthe amorphous state in general, glassy or molten. The mode! was based onstructural evidence from electron diffraction and dark field microscopy studiesof crystallizable and non-crystallizable polymers. The model contains twomajor elements as depicted in Figure 23. One element, which he called the grain(--30—100A), consists of an ordered domain of parallel chain segmentsbrought about primarily by intramolecular folding of parts of a long chainmolecule. The grains are generally connected to one another as in a string ofbeads especially for those molecules that are longer than can be containedwithin a single grain. Surrounding the grains are the second phase (-. l0—50A)consisting primarily of molecular chain segments in a truly disordered state,which may include low molecular weight molecules, chain ends or segments ofmolecules going from one grain to another.

A natural consequence of this model is the concentration of excess freevolume in the matrix phase. A detailed description of this model and its possiblerelation to lamellar crystallization, rubber elasticity, viscosity, etc. have beencovered in the same article.

86

MORPHOLOGY OF AMORPHOUS POLYMERS

Figure 23. Schematic representation of the fold-chain fringed-micellar grain model showing theordered domain (O.D.) the grain boundary layer (GB.) and the intergraio region (1G.)35.

Another "two-phase" model which needs to be mentioned is the one beingdeveloped by Pechhold.43 Starting with a nondefective (all trans conformations)bundle of molecules, by the introduction of more and more molecular defects(e.g. gauche conformations) into the system, Pechhold calculated from a simplethermodynamic treatment the mean radius of a meander curvature (disorderedregion). This radius came out to be about 50A. His model is reproduced in Figure24.

Figure 24. Pechhold's model for the amorphous molten state of polymers43.

87

G. S. Y. YEH

All four models have one major structural property in common. Themolecules or segments of molecules tend to assume approximate parallelalignment over large regions compared to repeat units. Another commonfeature is that in none of the four models is there any extended region (> 2—3) ofcrystalline-type trans conformations remaining in the amorphous state. In twoof the four (namely by Yeh and Pechhold) some definite regions of more highlydisordered chain segments are included; while in three of the four (namelyHosemann. Yeh and Pechhold) some folded regions are also suggested.

We hope that the presentation and a brief discussion of these models here (allof which. we are sure, will be modified to a certain extent when more experimen-tal evidence presents itself) may help to clarify the meaning of order (and/orstructure) in the amorphous state of polymers. This may be of particularimportance to those who wish to extend or improve these various models bycalculations or experiments.

REFERENCESP. H. Geil. Polymer Single Crystals, lnterscience, New York (1963).

2 A. Peterlin. J. Polymer Sci. 9. 61(1965).E. W. Fischer and R. Lorenz. Kolloid-Z. 189. 97 (1963).J. B. Jackson. P. J. Flory and R. Chiang. Trans. Faraday Soc. 59. 1906 (1963).T. Kawai and A. Keller. Phil. Mag. 8. 203 (1963).M. Takayanagi. K. Imada. A. Nagai. T. Tasumi and T. Matsuo. J. Polymer Sc!. C 16. 867(1967).T. Tasumi.T. Fukashima. K. Imada and M. Takayanagi.J. Macromol. Sc!. B1.459 (1967).

8 01-1. D. Keith, F. J. Padden. Jr.. and R. G. Vadimsky. Science. 150. 1026 (1965). "1-1. D.Keith. F. J. Padden.Jr.and R. G. Vadimsky.J. Polymer Sci. Al 267(1966). '1-1. D. Keith, F.J.Padden. Jr. and R. G. Vadimsky. J. Appl. Phvs. 37. 4027 (1966).G. S. Y. Yeh. Morphology ofAmorphous Polymers, to he published in CRC Critical Review inMacromolecular Science.

0 E. Guth. J. Polymer Sci. C 12, 89 (1966).° R. Blokland. Ph.D. Thesis, University of Rotterdam (1968).2 K. Nagai.J. Chem. P/is's. 47. 4690 (1967).

K. Nagai. J. Chem. Ph;'s. 49. 4212 (1968).A. N. Gent. Macromol. 2. 262 (1969).' R. E. Robertson. J. Pkvs. Chem. 69. 1575 (1965).

0 R. D. Heidenreich. W. M. 1-less and L. L. Ban. J. AppI. Cryst. I. 1 (1968).' S. Krimm. J. P/n's. Chem. 57. 22 (1953).A. Bjornhaug. 0. Ellefsen and B. A. Tonnesen. J. Polymer Sc!. 12, 621 (1954).

19 H. G. Kilian and K. Boueke. J. Polymer Sci. 58. 311(1962).20 K. Katada. Acta Cryst. 16. 290 (1963).21 A. Chapiro. Radiation Chemistry of Polymeric Systems, lnterscience. Wiley. New York

(1962).22 G. S. Y. Yeh. Order in amorphous polystyrenes as revealed by electron diffraction and

diffraction microscopy. J. Macromol. Sc!. B. In press.21 s L. Lambert. Ph.D. Thesis, University of Michigan (1970).24 T. G. F. Schoon. Brit. Polymer J. 2, 86 (1970).2 G. S. Y. Yeh and S. L. Lambert. Mechanism of spherulitic crystallization as deduced from

observations of partially crystallized glassy polystyrene. Submitted for publication in J.Macromol. Sci. B.

21, G. S. Y. Yeh and P. H. Geil. J. Macromol. Sci. B 1(2). 235 (1967).27 5• Carr, P. H. Geil and E. Baer. J. Macromol. Sc!. B2(1), 13 (1968).28 A. Siegmann and P. H. Geil. J. Macromol. Sc!. B4(2). 239 (1970).29 A. Ermolina. G. Markova and V. A. Kargin. Study of structural changes in polyethylene

terephihalale and polychlorotr(fluorethylene over a range of melting temperatures of crystals,AllUnion Symposium on Electronography. Moscow (1957).

88

MORPHOLOGY OF AMORPHOUS POLYMERS° J. J. Kiement. Ph.D. Thesis, Case Western Reserve University (1969).

W. Frank, H. Goddar and H. A. Stuart. J. Polymer Sci. B5, 711, (1967).32 G. S. Y. Yeh and P. H. Geil. J. Macromol. Sci. B1(2), 251 (1967).

G. L. Simard and B. E. Warren. J. Am. Chem. Soc. 58, 507 (1936).34 D. Luch. Ph.D. Thesis, University of Michigan (1971).

G. S. Y. Yeh. A structural model for the amorphous state of polymers:folded chain fringedmicellar grain model, J. Macromol. Set. B. In press.

36 L. L. Ban. Private communication.aJ R. Katz. Z. Physik. Chem. (Leipzig), A125, 321 (1927). bJ R. Katz. Trans. FaradaySoc., 32, 77 (1936).

38 V. A. Kargin, A. I. Kitaigorodskii and G. L. Slonimskii. Kolloidn. Zh. 19, 131(1957).a Bose. Physik. Z. 8, 513 (1907). "E. Bose. Physik.Z. 10, 230 (1909).° L. S. Ornstein. Z. Krist. 79, 10 (1931).

41 L. S. Ornstein and W. Kast. Trans. Faraday Soc. 29, 881 (1933).42 R. Hosemann. J. Polymer Set. C20, 1 (1967).U w Pechhold. IUPAC Preprints, 789 (1971).

89