Embed Size (px)
Citation preview
••• National LibraryolCanada
Bibliothèque nationaledu Canada
Acc,uisitions and Direction des acquisilions etBibliographic Services Branch des services bibliographiques
395 Welhnglon Streel 395. rue Wellingtononawa. Onl3no Onawa (Onlano)K1AON4 K1AON4
NOTICE AVIS
The quality of this microform isheavily dependent upon thequality of the original thesissubmitted for microfilming.Every effort has been made toensure the highest quality ofreproduction possible.
If pages are missing, contact theuniversity which granted thedegree.
Some pages may have indistinctprint especially if the originalpages were typed with a poortypewriter ribbon or if theuniversity sent us an inferiorphotocopy.
Reproduction in full or in part ofthis microform is governed bythe Canadian Copyright Act,R.S.C. 1970, c. C-30, andsubsequent amendments.
Canada
La qualité de cette microformedépend grandement de la qualitéde la thèse soumise aumicrofilmage. Nous avons toutfait pour assurer une qualitésupérieure de reproduction.
S'il manque des pages, veuillezcommuniquer avec l'universitéqui a conféré le grade.
La qualité d'impression decertaines pages peut laisser àdésirer, surtout si les pagesoriginales ont étédactylographiées à l'aide d'unruban usé ou si l'université nousa fait parvenir une photocopie dequalité inférieure.
La reproduction, même partielle,de cette microforme est soumiseà la Loi canadienne sur le droitd'auteur, SRC 1970, c. C-30, etses amendements subséquents.
•CRYSTALLIZATION AND MORPHOLOGY
OF OPTICALLY ACTIVE POLYETHERS
by
Kathy L. Singtield
A thesis submitted ta the Faculty ofGraduate Studies and Research
in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
Depanment of Chemistry
McGiIl University
Montreal, Quebec
Canada©Kallty L. Singfield
January, 1996
.+. National Libraryof Canada
Bibliothèque nationaledu Canada
Acquisitions and Direction des acquisitions etBibliographie Services Rr"."cll des services bibliographiques
395 Wellington StreetOttawa. OntarioK1A ON4
3~5. rue WellingtonOttawa (Onlano)K1AON4
The author has granted anirrevocable non·exclusive licenceallowing the National Library ofCanada to reproduce, loan,distribute or sell c"pies ofhis/her thesis by any means andin any form or format, makingthis thesis available to interestedpersons.
The author ~etains ownership ofthe copyright in his/her thesis.Neither the thesis nor substantialextracts from it may be printed orotherwise reproduced withouthis/her permission.
l'auteur a accordé une licenceirrévocab!9 et non exclusivepermettant à la Bibliothèquenationale du Canada dereproduire, prêter, distribuer ouvendre des copies de sa thèsede quelque manière et sousquelque forme que ce soit pourmettre des exemplaires de cettethèse à la disposition despersonnes intéressées.
L'auteur conserve la propriété dudroit d'auteur qui protège sathèse. Ni la thèse ni des extraitssubstantiels de celle-ci nedoivent être imprimés ouautrement reproduits sans sonautorisation.
ISBN 0-612-12487-8
Canada
•
•
•
Absrracr
ABSTRACT
The isothemlal crystallizaùon kinetics. spherulite morphology, and thennal
hchavior of the meh·crystallizcd opùcally acÙve R and S polyenanùomers, their hlends.
and the stereoblock fonn of poly(cpichlorohydrin) (PECH) and poly(propylene oxide)
(PPrO) have hcen invesùgated using polarized light microscopy (PLOM), atomic force
microscopy (AFM). and differcnùal scanning calorimetry (OSC). The novel opùcally pure
polyenantiomers of PECH were synthesized l'rom the opùcally acùve monomers using a
triethylaluminum·water (1:0.6) catalyst. which was also used in the polymerizaùon of the
racemic monomer to an isotacùc cryslalline stereoblock fraction. The opùcal!y pure
polyenantiomers and the stercoblock fonn cf PPrO were obtained by the quanùtaùve
dechlorinaùon reacùon of the corresponding PECH polymers, using LiAlH4• with
retention of contiguration of the main chain chiral centers.
The spherulite :"!dial growth rates of the melt-crystallized PECH equimolar
polyenantiomer blend are depressed relaùve to those of either opùcally pure components
over the range of crystallizaùon temperatures l'rom the glass transition temperature
(Tg =-26 oC) to the equilibrium melting temperature (TmO = 138 oC), which were
detennined to he the same for all of the PECH polymers. A further marked reducùon in
growth rates is recorded for the stereoblock polymer. Conversion of the PECH
polyenanùomers to PPrO results in an overall order of magnitude increase in the spherulite
radial growth rates. The growth rates of the PPrO stereoblock are only slightly depressed
relative to those of either opùcally pure polyenanùomer over the range of crystallizaùon
temperatures l'rom Tg (-65 oC) to Tm° (82 oC), which were detennined to he the same for
all of the PPrO polymers. An analysis of the linear spherulite radial growth rates of the
PECH polymers in terms of the Hoffman-Lauritzen treaunent gives evidence of a rougher
lamellar fold surface in the stereoblock polymer than in the polyenanùomers, but suggests
similar surface free energies among the polyenanùomers and stereoblock form of POO.
The mulùple melting behavior exhibited by the different forms of PECH is greatly
dependent on OSC heating rate and isothermal crystallizaùon ùme, for samples
crystallized at low temperatures and high temperatures. respecùvely. The hehavior is
demonstrated to he due to reorganization during the OSC heating scan in the fonner case,
and linked to the process of secondary crystallizaùon in the latter.
ii
•
•
•
Ahsrracr
The optically pure polyenantiomers. thcir equimolar hlcnd. and the slereohlm:k
forrn of PPrO ellhihit regularly handed spherulites. ohser\'ed using PLOM. In eontrast.
only the optically active polyenantiomers of PECH forrn handed spherulites. whereas the
cquimolar polyenantiomer hlend and the slereohlock display nonhanded. eoarser
spherulites. The hirefringent elltinction handing pattern of the PECH optically pure
polyenantiomer spherulites corresponds directly to the surface topography mapped hy
AFM: Regularly alternating concemric ridges and valleys indicate the edges and the fold
surfaces. respectively. of the radiating helicoidal larnellae. The direction or "sense" of an
apparent surface spiral pattern of a banded spherulite is directly dependent 0n the chiml
sense of the constituent polyenantiorner. It is suggested that the effecl~ of the hackhone
chirality are being transrnined to the level of the gross spherulite rnorphology.
On the basis of the observed differences in crysmllization kinetics and spherulite
rnorphology arnong the weil characterized PECH polyrners. a rnechanism of stereospccilic
segregation at the growth front is proposed.
iü
•Résumé
RÉSUMÉ
La cinétique de cristallisation isotherme, la morphologie sphérulitique, et le
comportement thermique des polyénantionomères optiquement actifs R et S cristallisés à
partir de l'état fondu, leurs mélanges, et la forme stéréo-séquencé de
poly(épichlorohydrine) (PECH) et de poly(oxypropylène) (PPrO) ont été investiguées à
l'aide de la microscopie à lumière polarisée (PLOM), la microscopie à force atomique
(AFM), et par calorimétrie différentielle à balayage (DSC). Les nouveaux
polyénantiomères pures et optiquement actifs de PECH ont été synthétisés à partir de
monomères optiquement actifs à l'aide d'un catalyseur de tryéthylaluminiumoeau (1:0,6).
Ce catalyseur a aussi été utilisé pour la polymérisation du monomère racémique. qui donne
la fraction contenant la forme stéréo-séquencé isotactique et crystalline. À l'aide du
LiAlH4• les énantiomères pures et optiquement actifs et le PPrO stéréo-séquencé ont été
obtenus par déchlorination quantitative des polymères de PECH, ces polymères gardent la
même configuration au niveau des carbones chiraux de la chaîne principale.
Les taux de croissanre radiale des sphérulites crystalisées à partir de l'état fondu,
pour un mélange équimolaire de polyénantiomères de PECH. sont réduits
comparativement à ceux des deux énantiomères pures sur l'intervalle de température de
crystallisation de la température de verre (Tg = -26 oC) à la température de fusion (Tm =
138 oC). Les taux de croissance radiale sont identiques pour tout les polymères de PECH.
La diminution du taux de croissance radiale est encore plus prononcée pour le polymère
stéréo-séquencé. La conversion des polyénantiomères de PECH en PPrO se traduit par
une augmentation du taux de croissance radiale des sphérulites. Le taux de croissance du
PPrO stéréo-séquencé est légèrement inférieur à ceux observés pour les deux
énantiomères pures optiquement actifs sur l'intervalle de crystallisation de Tg ( -65 OC) à
Tm (82 OC). qui ont été déterminés comme étant identiques pour les polymères de PPrO.
Une analyse des taux de croissance radial des sphérulites selon Hoffman-Lauritzen, pour
les polymères de PECH. met en évidence que la surface des replis lamellaires est moins
uniforme pour le polymère stéréo-séquencé. mais cette analyse suggère que l'énergie libre
de la surface est similaire pour les polyénantiomères de PPrO et le PPrO stéréo-séquencé.
La présence de multiples comportements de fusion pour le système basé sur le
PECH dépend du taux de chauffage utilisé en DSC pour les échantillons cristallisés à basse
température. et à haute température sur le temps de cristallisation isothermal. Dans le
iv
•
•
premier cas. à basse température lorsque la DSC est utilisée. la réorganisatilln durant le
balayage est responsable du phénomène. alors que dans le secllnd cas. un mécanisme de
cristallisation secondaire est responable de cet effet
Des sphérulites à bandes régulières sont observées. par la PLOM. pllnr les
polyénantiomères optiquement pures. leur mélange équimolaire. et la forme stéréll
séquencé du PPrO. Par contre, pour le PECH. seulement les polyénantillmères
optiquement actifs forment des sphérulites à bandes. alors que les mélanges équimlllaires
et la forme stéréo-séquencé du PECH produisent des sphérulites sans handes et mllins bien
délinies. Le patron d'extinction de la birefringence des bandes des sphérulites des
polyénantiomères de PECH optiquement pures correspond directement à la topologie de
la surface obtenue par AFM: Des crêtes et des vallés concentriques alternant
régulièrement indique. respectivement. les rebords et les replis des surfaces des lamelles
hélicoïdales irradiantes. La direction ou le "sens" du patron apparemment en spiml de la
surtàce pour les sphérulites à bandes dépend de l'orientation chirJJe du polyénantiomère
utilisé. Ceci suggère que la chiralité de la chaîne est transmise au niveau de la
morphologie générale de la sphérulite.
Basé sur les résultats obtenus lors des cinétiques de crystallisation. ainsi que les
morphologies observées pour les polymères de PECH. un mécanisme de ségrégation
stéréo-spécilique au niveau du front de croissance des sphérulites est proposé.
v
•
•
•
ACKNOWLEDGMENTS
It is my pleasure to express my sincere gratitude to Professor G. Ronald Brown,
for his help, supervision, and devotion to this work. 1 a1so acknowledge Dr. Brown for
having lirst introduced me to the subject of Polymer Chemistry with much enthusiasm, and
later, for sharing my own eagemess to explore new ideas throughout the growth and
development of this projecL
•ft has been my good fortune to learn from the generous advice and the wisdom of
Professor (Emeritus) Leon St.-Pierre, during my term at McGiII.
•1 am grateful to Dr. R. St. John Manley for his constructive criticism and helpful
suggestions from which this thesis has benelited greatly.
•1wish to thank Dr. Greg Johnston for helpful discussions conceming the synthesis
of optically active materials.
•1 am indebted to my friends and colleagues for their contributions to this project in
providing an amicable and stimulating environmenL In particular 1extend a special thanks
to Dr. Victor Caldas for the innumerable discussions and debates. 1 gratefully
acknowledge the efforts of Ms. Joy Klass for her contributions in performing many atomic
force microscopy investigations. For his assistance in attaining the X-ray diffraction
patterns, and for the invariably interesting discussions concerning the molecular
conformations of the polymers studied. 1 thank Mr. Die Saracovan. 1 aIso appreciate the
efforts and photographic skills of Mr. Neil Cameron in acquiring sorne of the photographs
contained in this thesis. 1 am grateful to Ms. Shanti Singh for her constant willingness to
listen and discuss new ideas. and to Ms. Cella Williams for her friendly encouragement
vi
•
•
•
AcknowledgmellCs
during the writing of this thesis. Many thanks are extended to Dr. Joon-Seop Kim for his
general assistance with the synthetic operations, the i::Onslruction and formatting of many
of the figures in the thesis, and for his fine camaraderie. 1 am grateful ta Ms. Antonella
Badia, for her skills and instruction on the atomic force microscope.
•For their involvement in the data collection, primarily in the polymer crystallization
Chapter, 1 wish to thank my summer students Ms. Norma Frangos, and in particular
Ms. Quynh Tan, who has continued to pursue her interest in the projecl.
•For helping 10 maintain the equipmenl used during my research, and in sorne cases
construct the apparatus used, 1 am graleful to Mr.'s Fred Kluck, George Kopp, Roland
Gaulin, and Rick Rossi.
•1 extend my appreciation to Dr. Françoise Sauriol, and Dr. Youlu Yu for their
skills and expertise in obtaining the NMR spectra included in Chapter Two.
•For performing the OPC analysis of the poly(epichlorohydrin) samples used in this
study 1 acknowledge Ms. Giselle Crone, and Dr. A. Rudin of the University of Waterloo
for his OPC analysis of the poly(propylene oxide) polymers.
•1 acknowledge the financial assistance of the Natural Science and Engineering
Research Council (NSERC) of Canada and Le Fonds pour La Formation de Chercheurs et
L'Aide à la Recherche (FCAR).
•Finally, my warmest wishes of appreciation are to my Family, and te my partner,
Dr. Paul Hemandez, for their love, patience and understanding throughout my studies.
Thankyou.
•••
vii
• TABLE OF CONTENTS
Abstract ----------------------------------------------------------------------------------- u
Résumé ------------------------------------------------------------------------------------ iv
Ackn0 wiedgments ---------------------------- ------------------------------------------- vi
Table 0 l' Conten ts ------------------------------------------------------------------------ viü
List of Symbols --------------------------------------------------------------------------- xv
List of Figures --------------------------------------------------------------------------- xix
List of Tables ----------------------------------------------------------------------------- xxvü
ChapterOne
GENERAL INTRODUCTION
1.1 Polymers: Macromolecules ------------------------------------------------- 1-1
1.2 Molecular Requirements for Crystallinity ------------------------------- 1-2
1.3 Crystallization of Polymers -------------------------------------------------- 1-5
1.4 Polymer Morphology ---------------------------------------------------------- 1-6
1.5 Optical Aetivity ------------------------------------------------------------------ 1-10
1.5.1 Plane Polarized Light ----------------------------------------------------- 1-10
1.5.2 Optically Active Molecules ----.------------------------------------------ 1-11
1.5.3 Stereoisomerism Nomenclature ---------------------------------------- 1-12
1.5.4 Optically Active Polymers --------------------------------------------- 1-14
1.5.4.1 Po1y(epich10rohydrin) ------------------------.----------------- 1-15
1.6 The Present Work -------------------------------------------------------------- 1-17
1.7 References ------------------------------------------------------------------------ 1-19
viü
Table (1 COlltellls
Chapter Two
POLYMER SYNTHESIS & CHARACTERIZATlON
2.1 Introduction: Ring· Qpening Polymerization ------------------------- 2-1
2.1.1 General --------------------------------------------------------------------- 2-1
2.1.2 Stereoregular Polyethers ------------------------------------------------- 2-~
2.1.3 Historical Perspective ---------------------------------------------------- 2-4
2.1.3.1 Coordination Catalysts ---------------------------------------- 2-4
2.\.3.2 Alkyl AIuminoxane Catalysts -------------------------------- 2-7
2.1.4 Mechanism of Alkyl AIuminoxane Catalyst Polymerization --------- 2-9
2.1.5 Chemical Modification of Poly(epichlorohydrin) --------------------- 2-11
2.1.6 The Present Work ------------------------------------------------------ 2-1 ~
2.2 Experimental -----------------------....--------..------....---------.-------.--- 2-14
2.2.1 Materials -----.-----.--.-------.--.-------•••-------.-.--------..-------.--- 2-14
2.2.2 Methods: General Overview ---.---------.-.---------.--------.------•.-- 2-14
2.2.3 Formation of EtJAl·O.6H20 ••••-----------------•.••----------------.---- 2-15
2.2.3.1 Apparatus ------•••••••----------.---------••---------.-----••--- 2-15
2.2.3.2 Dilution of Et.lAl in Hexane ---------.-.-.-------.------------ 2-17
2.2.3.3 EtJAI Reaction with Water ----.----------.-----------------.- 2-19
2.2.3.4 Storage of Catalyst Solution --------.-.---------------------- 2-19
2.2.4 Polymerization of Epichlorohydrin ---.-----------------------------.--- 2-211
2.2.5 Catalyst Removal -•••------------••-.--------•••-------_.-••------.-.----- 2-21
2.2.6 Separation of Atactic and Isotactic Poly(epichlorohydrin) -----.----- 2-21
2.2.7 Polymerization of Optically Pure Monomers ----••-------------------- 2-22
2.2.8 Dechlorination of Poly(epichlorohydrin) ------------------------------- 2-22
2.2.9 Preparation of Polyenantiomer Blends -_••-----------_._--------_.----- 2-24
2.2.9.1 Poly(epichlorohydrin) ---------------------.------.------------ 2-24
ix
Table ofConrenrs
2.2.9.2 Poly(propylene oxide) ----------------------------------------- 2-24
2.3 Polymer Characterization -------------------------------------------------- 2-24
2.3.1 Gel Permeation Chromatography (GPC) ------------------------------- 2-24
2.3.2 IlC Nuclear Magnetic Resonance SpeclroscoPy (IlC NMR) -------- 2-25
2.3.2.1 Poly(epichlorohydrln) ---------------------------------------- 2-25
2.3.2.2 Poly(propylene oxide) ----------------------------------------- 2-30
2.3.3 Polarimetry ---------------------------------------------------------------- 2-33
2.3.4 Fourier Transform Infrared SpeclroscoPy (FT-IR) ------------------- 2-34
2.3.5 Wide-Angle X-Ray (WAX) Diffraction -------------------------------- 2-36
2.3.6 Differentiai Scanning Calorimetry (OSC) ------------------------------ 2-41
2.3.6.1 Glass Transition Temperature (T.) --------------------------- 2-41
2.3.6.2 Equilibrium Melting Temperature (Tm0) --------------------- 2-44
2.4 Summary ------------------------------------------------------------------------- 2-47
2.5 References -------.-------------------------------------.---------------..------- 2-49
Chapter Three
THERMAL BEHAVIOR
3.1 Introduction: Muniple Melting Behavlor in Polymers -------------- 3-1
3.1.1 Historical Note ------------------------------------------------------...-- 3-3
3.1.1.1 Polyrner Single Crystals --••••-.---••----.-••••••--•••-••-. 3-3
3.1.1.2 Bulk Polyrners ••••--.---------.-------------.------.------- 3-3
3.1.1.3 Reorganization Model -------------------------.-----.------.- 3-6
3.1.2 The Present Study -----•••_-••••••••••--••----••••••-.-••••••----••---- 3·9
3.2 Experimental -.----------- --..-.------.------ ----.-- --- ---- 3-9
x
Table ofConrenrs
3.3 Results ---------------------------------------------------------------------------- 3-1 (}
3.3.1 Poly(epichlorohydrin) Cold-Crystallization ---------------------------- 3-lO
3.3.2 Multiple Melting in Poly(epichlorohydrin) ----------------------------- 3-lO
3.3.2.1 First Melt Thermograms -------------------------------------- 3-lO
3.3.2.2 Varying Heating Rate ExpcrimcnL~ -------------------------- 3-\3
3.3.2.3 Partial Heating Experiments ---------------------------------- 3-17
3.3.2.4 Varying Crystallization Temperature ExpcrimcnL~ --------- 3-19
3.3.2.5 Poly(propylene oxide) (PPrO) -------------------------------- .i-22
3.4 Discussion --------------------------------------------------- ..------------------ 3-26
3.4.1 Effect of Varying the Heating Rate ------------------------------------- 3-26
3.4.2 Effect of Partial Heating ------------------------------------------------- 3-29
3.4.3 Effect of Varying Crystallization Temperature ------------------------ 3-31
3.4.3.1 Secondary Crystallization ------------------------------------- 3-32
3.4.3.2 Comment on the Ratio of the Lower to Higher
Temperature Peaks ------------------------------------•••----- 3-36
3.4.3.3 Remark on the Conditions of the Hoffman-Weeks Plot .-- 3-37
3.4.4 Note on the Replacement of the Chlorine Atom in PECH ----------- 3-39
3.4.5 Concluding Comments on Mult'ple Melting ••••----------------------- 3-40
3.5 Summary and Conclusions .-------••••----------.--.-----------.----------- 3-40
3.6 References --_._-------------------------------------------------_.--------------- 3-42
Chapter Four
CRYSTALLIZATION KINETICS
4.1 Introduction
4.1.1 Stereocomplexation Behaviorin Polyenantiomer Blends -----------
4-1
4-1
xi
Table ofContents
•
•
•
4.1.2 Stereocomplexation Behavior in Products of Stereoselective
Polymerizati0 n ------------------------------------------------------------
4.1.3 Other Levels of 'Optical Compensation' -------------------------------
4.1.4 Optically Active Poly(epichlorohydrin) & Poly(propylene oxide) --
4.1.5 General Polymer Crystallization Kinetics ------------------------------
4.1.5.1 Growth Regimes -----------------------------------------------
4.1.6 The Present Study -------------------------------------------------------
4.2 Experimental --------------------------------------------------------------------
4.2.1 ~aterials --------..-----------..--------------------------------------------
4.2.2 Polarized Light Video ~icroscopy -------------------------------------
4.2.2.1 Apparatus ---------.-----------------•••••--.---••--••••-••----••
4.2.2.2 Sample Preparation ••----••••---------•••••---•••••••--.-•••-••
4.2.2.3 Spherulite Radial Growth Rate ~easurements .--•••-.-••--
4.2.2.4 ~easurementof Spherulite Band Period -.-•••••••--••- .
4.3 Results --.--.- - -- - .
4.3.1 General Spherulite ~orphology •••- - .
4.3.1.1 Poly(epichlorohydrin) ••-••••- - ••••••••••••••••••••••••
4.3.1.2 Poly(propylene oxide) ••••••••••••••••••-••••••••••••••••••••••
4.3.1.3 The Temperature Dependence of Band Period --.
4.3.2 Spherulite Radial Growth Rates --••••••••••••••••- ••••••••- •••••••••••
4.3.2.1 Poly(epichlorohydrin) •••- ••••••- ••- •••••••••••••••••••••-.
4.3.2.2 Poly(propylene oxide) ••••••••••••••- ••••••••- ••- •••- •••-.
4.4 Discussion -- -.- - -- -
4.4.1 Spherulite Radial Growth Rates: Poly(epichlorohydrin) -••-••••••-.-
4.4.1.1 Assignment of the Regime Coefficient (j) •••-•••••••••••••••
4.4.1.2 Lateral and Fold Surface Free Energy (erer.) ••••••••--••••••
4.4.2 Spherulite Radial GIOwth Rates: Poly(propylene oxide) .-----.-
4-3
4-4
4-6
4-7
4-8
4-11
4-11
4-11
4-12
4-12
4-12
4-13
4-13
4-13
4-13
4-14
4-14
4-21
4-27
4-27
4-34
4-38
4·38
4-39
4-40
4-41
xii
•Table ofContents
4.4.2.1 Assignment of the Regime Coeftïcient (j) ------------------- 4-42
4.4.3 Morpho1ogy and the Crystallizaüon Kineücs -------------------------- 4-45
4.5 Summary & Conclusions ---------------------------------------------------- 4-47
4.6 References ---------------------------------------------------------------------- 4-4g
Chapter Five
SPHERULITE MORPHOLOGY
5.1 Introduction -----------------------..---------------.-----------------------.----- S-I
5.1.1 Banding in Po1ymer Spherulites --.-------------------------------------- S-2
5.1.1.1 Current Molle1s ------------------------------------------------ S-3
5.1.2 Methods of Investigation --------•••----------.------------------------- 5-S
5.1.3 The Present Work -------••----------------------------------------------- S-S
5.2 Experimental ------------------------------------.------------------•••---------- 5-6
5.2.1 Materials ------.-.------------------•••----------••----------------.--.----- 5-6
5.2.2 Po1arized Light Opücal Microscopy (PLDM) -••-------------•••------ S-6
5.2.3 Reflectance Light Dptical Microscopy (RLDM) -------------.-------- 5-6
5.2.4 Atomic Force Microscopy (AFM) ----------------.----------------.... 5-7
5.2.4.1 Basic Operating Principles --------.-----------------.-.-.----- 5-7
5.2.4.2 AFM Instrument and Settings Used in the Present
Investigation -.-.----------••---------••••------------.--••--.. 5-1Il
5.2.4.3 Samp1e Preparation -.-----------------.------------.---.------ 5-11
5.2.4.4 Image Processing -.---------------.-.----------.--.-.---------- 5-11
5.3 Results and Discussion --.------.-----.----------------.----.-------------.- 5-12
5.3.1 On the Three Dimensional Nature of the Spherulites ---••------------ 5-12
5.3.2 The Appearance of a Surface Pattern --.--------------••-.--.----------- 5·18
xili
•Table ofContents
5.3.3 Dendriùc Growth --------------------------------------------------------- 5-27
5.3.4 On Lamellar Twisùng ---------------------------------------------------- 5-36
5.3.5 Asymmetry as a Simple Chiral Factor ---------------------------------- 5-40
5.4 Summary and Conclusions ------------------------------------------------- 5-41
5.5 References ----------------------------------------------------------------------- 5-43
ChapterSïx
CONCLUSIONS, CONTRiBUTIONS TO ORIGINAL
KNOWLEDGE, & IDEAS FOR CONTINUED RESEARCH
6.1 Conclusions & Contributions to Original Knowledge -------------- 6-1
6.2 Ideas for Continued Research --------------------------------------------- 6-6
xiv
•y
[a.]
13C
Â
a.
AFM
AlCl3
ho
CDCl3
• CHCl)
cl
d
Mf
DMSO
DSC
âT
DTA
E~Al
fF.T.
• Feel)
FT-IR
LIST GIF SYMBOLS
Lamellar thickening factor
Specific optical rotation angle
Carbon isotope
Angstrom (10-10 m)
Empirical value
Atomic force microscopy
Aluminum chloride.
Tbickness of the crystallizing stem
Deuterated chloroform
Chloroform
Density of the pure liquid or the concentration of a sample solution
Dextrorotatory rotation direction
Heat of fusion per volume of monomer UnilS
Dimethyl sulfoxide
Differentiai scanning calorimetry
Difference between the equilibrium melting temperature (Tm0) and the
isothermal crystallization temperature (T.,) or degree of undercooling
Differential thermal analysis
Triethylaluminum
Variation of the heat of fusion away from the meltng point
Fourier transformed
Ferric chloride
Fourier-transformation infrared spectroscopy
xv
List ofSymbols
• G Gibbs free cnergy
G Radial growth rate
Go Pre-exponenù'll term
GPC Gel permealion chromalography
H Enthalpy
h Hour
H-L Hoffinan-Lauritzen
H-W Hoffinan-Weeks
HPLC High performance liquid chromalography
i-PP Isotacùc polypropylene,
i-PRSECH Isotaclic fraction of poly(R,S-epichlorohydrin)
i-PRSPrO Isotactic fracùon of poly(R.S-propylene oxide)
i-PS L~otaclic polystyrene
• j Regime cœflicient
K Ahsolute temperature
k Bolt7.mann's constant
Kg Nucleation constant
KOH Potassium hydroxide
1 Levorotatory rotation direction
1 Path length of the sample cell for optical rotation angle measurement
LiAlH4 Lithium aluminum hydride
Mn Numher average molecular weight
Mw Weight average molecular weight
nm Nanometer (10"9 m)
NMR Nuclear magnetic resonance
PECH Poly(epichlorohydrin)
PEEK Poly(ether ester ketone) [= poly(aryletheretherketone)]
xvi
• PEO
PET
PHB
PLOM
ppm
PPrO
PRECH
PRPrO
PSECH
PSPrO
PVF2
R
R
r2
RLOM
rpm
RTD
S
S
0'
SAXS
0'.
T~
List of Symbols
Poly(ethylene oxide)
Poly(ethylene tetephthalate)
Poly(~-hydroxyhUlyrate)
Polarized light optical microscopy
Part pet million
Poly(propylene oxide)
Poly(R-cpichlorohydrin)
Poly(R-propylene oxide)
Poly(S-epichlorohydrin)
Poly(S-propylene oxide)
Poly(vinylidenc l1uoride)
Recrus absolute configuration; clockwise direction
Gas constant
Linear least-square correlation coefficient
Rel1ectance light optical microscopy
Revolutions per minute
Resistance temperature detector
Sinisrer absolute configuration; a counter-clockwise direction
Entropy
Lateral surface interfacial free energies
Small-angle X-ray scanering
Fold surface interfacial free energy
Hypothetical temperature at which molecular motion associated with
viscous l10w ceases
T. Crystallization temperature
Tg Glass transition temperaLUre
TIlF Tetrahydrofuran
xvii
• TiCl)
TiCl4
Tm
TCm
Trou
u*
WAX
WAXS
List ofSymbols
Titanium(III) chloride
Titanium(IV) chloride
Observed melting temperalUre
Equilibrium melting temperature
Crystallization temperature of maximum spherulite radial growth rate
Activation energy for transport of crystallizable segments through the melt
to the site of crystallization
Wide-angle X-ray
Wide-angle X-ray scattering
xviii
•Figure 1.1
Figure 1.2
Figure 1.3
LIST OF FIGURES
Sections of a poly(epichlorohydrin) chain showing (a) rcgioddcCIS. \-4
(b) isotacticity, and (c) syndiotacticity.
An illustration depicting the arrangement of polymer chains within 1-9
a lamella growing radially in a typical spherulite.
Isotactic chain configuration of poly(epichlorohydin) and 1-\3
poly(propylene made) demonstrating the absolute conligurations of
the chiral carbons on each chain.
•
•
Figure 1-4 Computer modeled lowest energy conformation of (a) an upward 1-16
poly(R-epichlorohydrin) helix, (b) a downward poly(R
epichlorohyilrin) helix, and (c) an upward poly(S-epichlorohydrin)
helix.
Figure 2.1 Semi-log plot showing dechlorination of poly(R.S-epichlorohydrin) 2-12
elastomer with LiAlH4 (2.5 mole/mole Cl) in THF at 50 oC (taken
from reference 28).
Figure 2.2 Schematic diagram of the polymerization apparatus. 2-16
Figure 2.3 !JC NMR spectra of the backbone carbons of (a) the crude 2-26
unfractionated PRSECH after one cold acetone wash, in CDCl) al
50 oC and (b) the atactic PRSECH cold acelone soluble fraction in
CDCl) at 50 oC, showing backbone carbon region only.
Figure 2.4 !JC NMR spectrum and expanded methylene region of i-PRSECH 2-28
fraction in deuterated DMSO at 50 oC.
xix
Figure 2.5
Figure 2.6
Figure 2.7
Figure 2.8
Figure 2.9
List ofFigures
I3C NMR spectrum and expanded Methylene region of PRECH in 2-29
deuterated OMSO at 50 oC.
I3C NMR spectrum of i-PRSPrO in deuterated chlorofonn at room 2-31
temperature.
I3C NMR spectrum of the product of the incomplete dechlorination 2-32
reaction of LiAlH4 on PSECH, in CDCI) at room temperature:
(a) PSECH and PRPrO backbone carbon region, (b) PSECH
pendent carbon region, and (c) PRPrO pendent carbon region.
Fr-IR spectra of (a) the equimolar blend of PSECH and PRECH 2-35
polyenantiomers and (b) i-PRSPrO.
Wide-angle X-ray powder diffraction patterns of (a) i-PRSECH, 2-38
(b) PRECH, and (c) the equimolar polyenantiomer blend.
Diffraction spots are due to the calibration standard (exposure time
= 2-4 h).
Figure 2.10 Wide-angle X-ray tiber diffraction patterns of (a) PRECH and 2-39
(b) i-PRSECH (exposure time ::18 h).
Figure 2.11 Wide-angle X-ray powder diffraction pattern of i-PRSPrO 2-40
(exposure time =4h).
Figure 2.12 OSC thennograms of melt-quenched (a) PRECH, (b) PSECH, (c) 2-42
the equimolar polyenantiomer blend, and (d) i-PRSECH showing
the T. (all heating rates =10 oC/min).
Figure 2.13 OSC thennogram of melt-quenched i-PRSPrO showing the T. 2-43(heating rate =20 oC/min).
Figure 2.14 Typical OSC heating thermograms used in the construction of the 2-45
Hoffman-Weeks plots of {a) PSECH isothennally crystallized for
24 h and (b) PSPrO isothc:rmally crystallized for 15 h, both at an
undercooling of 30 oC, and heated at 20 oC/min.
xx
•List of Figures
Figure 2.15 Hoffman-Weeks plot of the observed melting temperature as a 2-46
funcùon of crystallization temperature for optically acùve PRECH
and PSECH (e). the equimolar polyenantiomer blend (.). and the
stereoblock i-PRSECH (&). The hollow symbols represent the
corresponding lower melting endotherm maxima.
Figure 2.16 Hoffman-Weeks plot of observed melùng temperature as a 2-48
function of crystallization temperature for i-PRSPrO.
Figure 3.1 OSC cooling thermogram of (a) PSECH (scanning rate = ZO 3-11
oC/min). (b) heaùng thermograms of PSECH. and (c) i-PRSECH
both quench-cooled l'rom the melt (heaùng rates =10 oC/min).
Figure 3.2 OSC frrsl-mel1 thermograms of (a) the as-fractionated i-PRSECH. 3-12
(b) the as-polymerized PRECH. and (c) the as-polymerized
PSECH. all heated at a rate of 20 OC/min. In (d) the as
polymerlzed PSECH is heated al a rate of 5 oC/min.
Figure 3.3 OSC thermograms of (a) PSECH and (b) the equimolar 3-14
polyenanùomer blend. isothermally crystallized al 90 oC for 1 h and
heated al varying rates.
Figure 3.4 The 10wer (e) and higher (.) peak lemperalures as a funclion of 3-15
OSC heaùng raIe for PSECH isothermally crystallized al 90 oC for
1 h. Hollow symbols =single peak temperalure.
Figure 3.5 OSC thermograms of PRECH isothermally crystallized al 90 oC for 3-18
1 h and (a) heated al 5 oc/min, (b) heated al 5 oC/min unùl high
temperalure peak, (c) same as (b) and immediately quench-cooled
10 90 oC and subsequently heated al 5 oC/min, and (d) same as (h)
and held al the high temperalure peak for 56 min, quench-cooled 10
90 oC, and heated al 5 oC/min.
Figure 3.6 OSC thermograrn of PRECH isothermally crystallized al 90 oC for 3-20
1 h and heated al a 5 oC/min unù1110 oC. held isothermal 42 min.
and then resumed heaùng al 5 oC/min.
xxi
• Figure 3.7
List ofFigures
OSC thennograms of (a) PRECH, (b) the equimolar 3-21
polyenantiomer blend, and (c) i-PRSECH samples isothennally
crystallized at Tc for ca. 24 h and subsequently heated at 20 oC/min.
FigLre 3.8 OSC thennograms of (a) PRPrO cooled from the melt at 20 3-23
oC/min, (b) PRPrO, (c) PSPrO, and (d) i-PRSPrO quench-cooled
from the melt and heated at a rate of JO oC/min.
Figure 3.9 OSC thennograms of i-PRSPrO isothennally crystallized at Tc for 3-25
20 min, quench-cooled to -25 oC, and heated at a rate of 20
oC/min.
Figure 3.10 Schematic illustration of the Rim and RuntZ! model of 3-27
reorganization in polymers during the thennal analysis heating scan.
Figure 3.1 1 OSC heating thennogram of the equimolar blend sample 3-33
isothennally crystallized at 110°C for 24 h and heated at (a) 20
oC/min and (b) 5 oC/min. (both thennograms calibrated with
indium at 20 oC/min.)
Figure 3.12 (a) DSC thennogram of PRECH isothennally crystallized at 90 oC 3-34
for different periods of time and heated at 20 oC/min and (b) plot of
partial heat of fusion of the lower and higher temperature peaks in
(a) as a function of crystallization time.
Figure 4.1 lllustration of (a) a spherulite radial growth rate curve and (b) a 4-8
theoretical Hoffman-Lauritzen plot showing growth regimes l, II
andm.
Figure 4.2 Polarized light optical photomicrographs of spherulites of (a) 4-15
optically active PRECH, (b) optically active PSECH, (c) the
stereoblock i-PRSECH, and (d) the equimolar blend of the
polyenantiomers.
xxii
• Figure 4.3
List of Figures
Polarized light optical photomicrographs of dendrites of (a) the 4-19
equimolar blend of PRECH and PSECH polyenantiomers: (h) and
(c) the stereoblock i-PRSECH, all melt-crystallized al 80 oC in min
sarnple sections.
Figure 4.4 Polarized light optical photomicrograph of a physical mixture of 4·20
PSECH and PRECH polyenantiomers mclted on me same glass
cover slip.
Figure 4.5 Polarized light optical photomicrographs of isomermally melt- 4-22
crystallized spherulites of (a) optically active PSPrO (Tc = 40 OC).
(b) PRPrO (Tc = 31°C). (c) me stereoblock i-PRSPrO (Tc =35 OC), and (d) the equimolar blend of PSPrO and PRPrO
polyenantiomers (Tc =32 oC).
Figure 4.6 (a) Relationship between the spherulite band period and 4-26
undercooling for me PRECH and PSECH polyenamiomers (O). the
95:5 blend of the polyenantiomers (0), and the 70:30 blend of the
polyenantiomers (â), and i·PRSPrO (e).
Figure 4.7 Polarized light optical photomierograph of a PRPrO step growth 4-28
spherulite showing the influence of temperature on the spherulitc
band period. The specifie crystal1ization temperatures are noted in
the tex!.
Figure 4.8 Plot of the spherulite radial growth rates dependenee on isothermal 4-29
crystallization temperature for the 95:5 (V), the 70:30 (0), and the
50:50 (0) molar blend of PSECH and PRECH polyenantiomers.
Figure 4.9 Plot of the spherulite linear radial growth rates dependenee on 4-30
isomermal crystal1ization temperature for PSECH and PRECH
polyenantiomers (e), the equimolar polyenantiomer blend (0), and
the stereoblock i-PRSECH (.).
•List ofFigures
Figure 4.10 Hoffman-Lauritzen plots for PRECH and PSECH (e). the 4-32
equimolar polyenantiomer blend (0). and the stereoblock
i-PRSECH (.).
Figure 4.11 (a) Optical photomicrograph of a typical spherulite of the 4-34
equimolar blend of PRPrO and PSPrO polyenantiomers showing
pleat fonnation and (b) The effect of the non-unifonn border of the
spherulite in (a) on the measurement of the spherulite radial growth
rate. (Tc =31°C)
Figure 4.12 Spherulite radial growth rate dependence on undercooling of 4-35
PRPrO. PSPrO (e), and i-PRSPrO (0), PRECH. PSECH (.), and
i-PRSECH (0) shwing the effect of dechlorination of PECH to
POO on crystallization kinetics.
Figure 4.13 Hoffman-Lauritzen plots for (a) PRPrO and PSPrO 4-37
polyenantiomers and (b) the stereoblock i-PRSPrO.
Figure 5.1
Figure 5.2
Figure 5.3
Figure 5.4
A schematic representation of a lamella possessing a series of 5-3
isochiral screw dislocations along its length (taken from reference
27).
Schematic illustration of the IWO operating regimes of the AFM, (a) 5-S
contact and (b) non-contact (taken from reference 3S).
Schematic illustration of the AFM instrument operating in the 5-9
contact operating regime (taken from reference 3S).
Reflectance optical micrographs of the unrestrained melt- 5-13
crystallized spherulitic film crystallized at SO oC of (a) the optically
pure polyenantiomer, PSECH, and (b) the equimolar blend.
xxiv
Lisr of Figures
• Figure 5.5 (a) Three dimensional AFM image of a quadrant of an optically S-IS
pure PSECH spherulite surface crystallized unrestrained at 80 oC.scanned in the constant force mode. (h) the corresponding
polarized light optical micrograph. (c) the mdial line inlCnsily
profile. showing the band periodicity. of the handed spherulitc
section shown in (b).
Figure 5.6 Low magnification AFM image scanned with a constant height of S-17
(a) PSECH spherulite and (b) PRECH spherulite. hoth melt-
crystal1ized unrestrained at 75 oC.
Figure 5.7 Higher magnification AFM images of the corresponding spherulitcs S-19
in Figure 5.6 scanned with a constant force of (a) PSECH. (b)
PRECH. (c) the high-pass filtered image of (a). and (d) the high-
pass filtered image of (b). Note the unique direction of inclination
of the lamellar edges in each of the isochiral spherulite surface
images, (e) The high-pass filtered image of (c) and (f) is the high-
pass filtered image of (d). The second filter process removes the
large banding structural features and leaves the line lamellar edge
structures.
Figure 5.8 Reflectance light optical micrographs of the crude surface of melt- 5-22
crystallized (a) PSECH and (h) PRECH spherulites. Tc =75 oC.
Figure 5.9 A schematic reconstruction of the computer-generated spherulite 5-26
surface of polyethylene. showing the C-shaped lamellar profiles.
The direction of the Cs denotes the handedness of the lamellae, and
the spherulite (taken l'rom reference 10).
Figure 5.10 AFM image of the unrestrained melt-crystal1ized surface of the 5-28
equimolar blend spherulite crystal1ized at 80 oC scanned with a
constant height
xxv
•
•
List ofFigures
Figure 5.11 High magnificaùon AFM images of the unrestrained surface of the 5-29
melt-crystallized spherulite at 80 oC scanned with a constant height
of (a) PSECH and (b) the equimolar blend.
Figure 5.12 Reflectance opùcal micrographs of the melt-crystallized dendriùc 5-31
structures grown in thin sections of the film at 80 oC of (a) PSECH
and (b) the equimolar b1end.
Figure 5.13 Low magnification AFM images of the dendritic structures 5-34
depicted in Figure 5-7 of (a) the optically pure po1yenantiomer
PSECH and (h) the equimolar b1end; and their respective higher
AFM images (c) and (d). AlI of the images were collected using
constant height mode.
Figure 5.14 A theoretical illustration of a cellulosic suspension showing a chiral 5-37
nematic phase behavior (taken from reference 46).
Figure 5.15 A schematic representation of a radiating helicoidal1amella with the 5-39
chain direction maintaining an orthogonal re1ationship with the top
and bottom 1amellar fo1d surfaces (taken from reference 47).
Figure 6.1
Figure 6.2
Figure 6.3
(a) Low magnification AFM image of PRPrO scanned in constant 6-10
height mode; (b) Po1arized light photomicrograph of i-PRSPrO
taken with a 1I4Â. wave plate.
Digitized video images of dendritic structures of PRPrO formed in 6-12
very thin sections of the me1t.
Reflectance light photomicrographs of PRECH spherulites. 6-13
showing the circumferential fractures. Details in the text.
xxvi
• LIST OF TABLES
Table 2.1 Optical Rotation Angles ofPoly(propylene oxide} Sarnples 2-33
Table 3.1 Heats of Fusion of PECH Crystallized at 90 oC for Different Times 3-16
Table 4.1 Best-fit Estimates of the Growth Rate Parameters for PECH 4-33
•
•
Table 4.2 Best-fit Estimates of the Growth Rate Pararneters for PPrO
Table 4.3 Estimates of 0'0'. for AlI Possible Values ofj for PECH
Table 4.4 Estimates of 0'0'. for AlI Possible Values ofj for PPrO
4-:;6
4-39
4-42
xxvü
•GENERAL
INTRODUCTION
ChapterOne
1.1 Polymers: Macromolecules
A polymer il' a large molecule, or macromolecule, in the form of a long chain, the
length of which is enormously large compared to its cross section (ca. typically 10 ()(){) to
1 as compared with 100 to 1 for a typical strand of spaghetti pasta).\ The long polymer
chain is made up of a large number of smaller units of itself, Le., it is self-similar. The
smallest of these repeat units is called a monomer. Through a control1ed chemical
reaction, appropriately initiated, the monomers are polymerized, or added in a sequential
fashion to each other forming the covalently-llnked polymer chain.
Sorne polymers possess the ability to crystallize. Due to their characteristic length
and self-similar nature, there is an obvious requirement of chain regularity for such a
cooperative, ordering process to occur among polymer molecules. When crystallizable
non-polymerie molecules are cooled to below their melting temperature (undercooled>,
crystallization generally results in the development of an essentially 100 % crysta\line solid
(neglecting small crystal defects). By contrast, the crystallization of polymers never
results in a 100 % crysta\line material. Polymer solids therefore possess a degree of
crystallinity. The chains, and/or chain segments which are excluded from the crysta\line
1-1
•
•
•
Genera/lnrroduction
regions of the rnaterial contribUle to the amorphous content of the rnaterial. The
arnorphous portions of this semicrystalline polymer are often refetred to as rejected
material, since it does not crysta11ize, at least not initially. With sorne crystallizable
polymers, crysta11ization can he suppressed C'.ompletely and rendered arnorphous in the
solid state by quickly cooling (quench-cooling) the polymer from ils arnorphous melt
state. Clearly, a non-crystallizable polymer always remains arnorphous upon cooling from
the melt state.
1.2 Molecular Requirements for Crystallinity
illtimately, the physical properties of polymeric rnaterials are strongly affected by
the degree of crysta11inity of the final solid state. At the fundarnentallevel, this depends on
the capability of the polymer chains to mutually align themselves along their chain axes in
well defined crystallattice sites. Chain regularity is therefore a requirement for polymer
crystallinity, and can he divided into three categories: (i) chemical regularity,
(ü) regioregu/ariry (sometimes called geometrical regularity), and (iü) tacticiry (sometimes
refetred to as spatial or stereoregularity).2
(i) The greatest level of cheoùcal regularity is achieved in a homopo/ymer, of
which the simplest one is polyethylene, composed only of [-C~-C~-] repeat units.
When the constituent monomer uoits are not identical the degree of cheoùcal regularity
lessens. Thus copolymers, which contain more than one type of monomer, are usually less
crystalline than the corresponding crystallizable homopolymer in the solid state. It follows
that a crystallizable copolymer composed of regularly altemating sequences of the
different monomers (block copolymer) is more crystalline in the final solid state !han the
extreme case ofa random distribution of different monomers (random copolymer).
1-2
•
•
General Introducrion
(ii) The polymer used in the studies presented in this thesis, poly(epichlorohydrin)
(PECH) can provide an example in Figure 1.1 (a) of the second crystallinity requirement
listed abovc. The section of the chain contained in Figure 1.1(a) is regioirregular. Il can
he seen that the repeat unit in PECH is nol symmetric: there is a distinction belWeen a so
called head-ro-rail attachment of monomers, and a head-ro-head or rail-ro-rail linkage.
The requirement of chain regularity is satisfied during polymerization, and depends on the
nature of the compound which initiates the polymerization {iniriaror or caralysr) and the
monomer itself. Points of irregular monomer addition relative to the growing chain are
called regio defect sites.
(iii) The spatial regularity of a polymer chain can he demonstrated with two
configurations of the planar zigzag (extended chain) conformation of PECH in Figure
l.l (b) and (c). Unlike polyethylene, PECH contains regular side groups, or pendent
groups, which in this case contain the chlorine atom of each repeat unit. The tacticity of
the polymer descrihes the regular placement and conjigurarional arrangements (flJœd,
relative positions) of these pendent groups along the polymer backbone, and thus
describes the polymer microstructure. There are three levels of tacricity:3 (i) In an
isotacric polymer ail of the repeat units have the same configuration (this usually implies
that in a vinylic polymer ail pendent groups are on the same side of the chain in a planar
zigzag conformation); (ü) neighboring repeat units in a syndiotactic polymer have
altemating configurations; (iii) an aracric polymer bas a random arrangement of
configurations. As can he seen in Figure 1.1, PECH contains an odd numher of backbone
atoms per repeat unit so that for an isotactic configuration, the pendent groups are
arranged on altemating sides (jf the polymer chain [Figure 1.1(b»); consequently, in a
syndiotactic configuration they are on the same side of the polymer chain [Figure 1.1(c»).
It can also he noted from Figure 1.1 that in every repeat unit in PECH there exists
a carbon atom which is bonded ta four different ligands. Such an asymmetric carbon atom
is termed a chiral carbon. By definition, the ligands can he bonded to the chiral atom in
1-3
•(a)
Tail-to-Head Tai\-to-Tai\
General Introduction
•
•
-O-I-C~-O-IH-CHZ-CH2IH-O-
CHzC\ CHzC\ CHzC\
(b) isotactic
(c) syndiotactic
Figure 1.1 Sections of a po\y(epichlorohydrin) chain showing (a) regio defects,
(b) isotaeticity, and (c) syndiotaeticity.
1-4
•
•
General Introduction
one of two possible enantiomeric configurations which are the non-superimposable mirror
images of each other.4 The subject of chiral carbon centers and how they accouni for the
ohserved optical activity in asymmetric molecules will be treated in greater detail later in
this Introduction. For now, il is important to note that in the case of polymers which
possess true chiral centers in the backbone, the crystallinity is dependent on the regularity
of thc type and spatial distribution of the chiral centers along the chain, Le., the tacticity
truly describes the stereoregularity in this case. Crystallinity is maximized when the
number of stereo defect sites (enantiomeric neighbors) is minimized. A copolymer which
contains stereo defect sites adjoining long sequences of isochiral carbon centers, is called
a stereoblock polymer. The longer the isochiral block, or isochiral stereosequellce, the
greater the capability of a stereoblock polymer to crystallize.
1.3 Crystallization of Polymers
At temperatures above the melting temperature of the polymer, the chain exists as
a random coil, and changes its conformation continuously. As the temperature is lowered
10 below the melting temperature however, the available volume to the polymer chain
decreases. The crystallizable polymer chain adopts one of a few basic conformations:
planar zigzag (as shown in Figure 1.1), or one of many helical-type conformations,
depending on the chemical and physical nature of the chain. In order for long chain
macromolecules to spontaneously form ordered structures upon undercooling from a
maximum disorder melt state, there must he a thermodynarnic driving force.S The Gibbs
free energy (G) of any system is related to the enthalpy (H) and the entropy (S) by the
equation,
G=H-TS. (1.1)
1-5
•General Introduction
where T is the thennodynamic temperalUre. In order for crystallizaùon to proceed. the
large negaùve entropy change which accompanies polymer crystallizaùon from the mc11 is
offset by a large decrease in enthalpy. A S'late of minimum free energy (G) will he allained
and thus crystalIizaùon will he thennodyna!nically favored when the decrease in enthalpy
(Mi) is greater than the product of the crystallizaùon temperature ùmes the decrease in
entropy (Tl1S).
For small molecules, crystalIization occurs rapidly at temperatures very close to
the melting temperature, and therefore the condiùons are very close to equilibrium. By
contrast, the crystalIization of polymer molecules reaches a substanùal rate only at
temperatures weIl helow the melting temperature. and thus far removed from equilibrium
conditions. Thus, polymer crystalIization tends to he controlled by kineùcs and the rate at
which the crystals nucleate and grow is very important in detennining the overall
crystallinity of the polymer solid.S
1.4 Polymer Morphology
The morphology of a polymer solid describes the size and shape and the mutual
positions of its constituent parts. Crystal1ine polymers exhibit a "hierarchical morphology"
so that there is a long-range order which spans a large range of dimensions.2 Chains in a
planar or helical confonnation, pack along their length inta higher order structures called
lamellae (crystallites). The basic repeat unit of the crystalIite is called the unit cell,
through which segments of typically 1 to 4 polymer chains pass.
The study of single lamellae fonned from a dilute polymer solution at temperatures
close to the polymer melting point pennits the study of a highly crystalIine example of
these fundamental structures. In the 1950's, a startling discovery was made in the study of
polyethylene single lamellae crystallized from dilute solution.6-8 It was found that the
1-6
•General Introduction
polymer chain axes of the polyethylene molecules. which in the fully extended form
exceeded lengths of 10 (){)() À. were aligned along the short dimension of the lamella. of
the order of 100 À thick. For the containment of polymer chains in this manner, it is a
necessary requirement that an incorporated chain must fold back on itself and re-enter the
lamellar crystal multiple times. thus creating fold surfaces on the top and bottom of the
lamella. perpendicular to the chain axes within.6 The lamellae which develop during
kinetically-controlled melt-crystal1ization at temperatures far l'rom the melting
temperature, do not reach the equilibrium form and possess a degree of disorder at their
laid sun·aces. Chain loops and loose ends (molecular cilia) contribute to the
intercrystallite and amorphous material which can hecome mobile at temperatures well
helow the melting temperature. Furthermore, sorne polymer chains may he associated
with more than one lamella. The non-crystal1ized chain segment which bridges the gap
hetween two lamellae (a rie molecule) contributes also to the non-crystalline regions of the
polymer solid.9 Thus, the crystal1ization of polymers l'rom the melt at appreciable rates
results in the formation of metastable crystal1ites.IO
In the hierarchy of the polymer solid state architecture, the lamellae are the
building blocks of the crystal1ite aggregates which develop during the melt-crystallization
of polymers. The dominant form of these aggregates is the spherulite. In addition to
polymers, spherulites are characteristic morphological forms of a broad spectrum of
substances including minerals ll and many small molecule substances.12 The growth of a
polymer spherulite can he initiated by random fluctuations in the density of a pure polymer
melt (homogeneous nucieation)13 or, as is more often the case, by the presence of non
polymeric nucleating surfaces present as irnpurities (heterogeneous nucieation).13
Spherulites display a spherical symmetry, arising l'rom the constant radial growth
in aIl directions of the component lamellae. The spherical symmetry does not however,
extend to the very center of the spherulite, where an observed sheaf-like structure is
helieved to he incipient to the mature spheruli!e.14 The opposite ends of the sheaf-like
1-7
General Inrroduction
structure meet and a spherical symmetric growth front forms and continues throughout
growth.15 The lamellae continually undergo a mechanism of non-crystallogrJphic
branching and fanning-out at non-crystallographic angles, thus providing uniform spacc
fIlling of the spherulite.14 On the other hand, iniLiated growth in ail directions from a
central nucleating entity, with the radiating lamellae undergoing crystallographic
branching, is indicative of growth of a dendritic structure. 15
The size of polymer spherulites typically ranges from 0.5 10 200 !lm and therefore
they are readily observable under an optical microscope. In general. polymer samples are
usually prepared between two microscope slides, and thus are examined in a thin film.
Upon cooling a sample from the mel!, this restrained preparation essentially limits the
spherulite growth to a flat disc instead of a sphere. As a result. the spherulite exhibil~ a
circular outline which is maintained throughout growth until impingement with another
spherulite entity. Straight impingement lines mark the boundary between two spherulites
possessing the same radial growth rate. Samples can a1.'0 he prepared with an
unrestrained top surface, however the development of a full sphere is restricted by the
eventual exhaustion of crystallizable material.
ln Figure 1.2 a typical polymer spherulite is presented. as viewed under a polarized
light microscope. with an illustration of the constituent lamellar building blocks extending
radially outward and showing the internai organization of the folded chain.~. Essentially.
lamellar growth is greatest in the spherulite radial direction. The chain stems are laid
down onto the lamellar growth face through a process of secondary nucleation and lateral
surface spreading across the crystal growth face.16 These growth mechanisms are
addressed specifically in Chapter Four.
It can he seen from Figure 1.2 that the polymer chains are oriented perpendicular
to the spherulite radial growth direction of the \amella in which they are contained. This
arrangement creates an optical anisotropy within the lamellae. That i~ to say that the
refractive index in the tangential direction is different to that in the perpendicular radial
1-8
G
•
•
..
Figure 1.2
Genera/Introduction
Jstem
An illustration depicting the arrangement of po\ymer ch:1Ïns within a lamella
growing radially in a typical spherulite.
\-9
•
•
•
General/mmdl/crion
direcùon. n.us spherulites are descrihed as heing birefringent. Due 10 the symmetrical
arrangement of the lamellae within the polymer spherulite. such optical anisotropy i~
manifested in the appearance of the characteristic spherulite Maltese cross ohserved in
polarized light under crossed polars15 (polari7..ed Iight is discussed in the following
section). Many polymer spherulites a1so exhihit a concentric hirefringent pattern in
addiùon to the Maltese extincticn cro.s. In these banded spherulires. it has hcen shown
that there is a periodic twisting of the crystallographic orientation ahout the lamellar axis
of radial growth.17•19
1.5 Optical Activity
1.5.1 Plane Polarized Light
The phenomenal property of light, or electromagnetic radiation, is thll mutual
orthogonal relationship which exislS among the componllnt oscillaùng ll111CtriC and
magneùc fields. and the direcùon of heam propagation. By considering the direction of
the electric vibrations. light can he depicted as a transverse wave with a cross sectional
pattern. Nonnal, unpolarized light is described as containing a multitude of directions of
electric vibrations with no particular net orientation. A 'snapshot' of a cross section of thll
oscillating electric field in normal unpolarized light at any given lime might he depictlld as
a vertical line (a cross section of an oscillation in the y-z plane ), or a horizontal Iinll (a
cross section of an oscillation in the x-z plane) in any of an infinite numher of directions in
space. In addition, the snapshot may appear circular, indicative of the cross section of thll
helical path traced in space by the electric field oscillating around any one axis. When
light is plane polarized the transverse vibrations are limited to sirnply one plane. Any
form of polarized light can he converted (polarized) to any other fonn with 100 %
1-10
•
•
General Introduction
efficiency.20 By comparison. sound waves ace longitudinal. instead of transverse. and
consequently cannot be polarized and have no sectional pattern (except for a point).
1.5.2 Optically Active Molecules
An optically active substance is one which rotates the plane of polarized light
through the interaction with the chacged particles of the molecules.20 That is to say: when
polarized light, with its transverse electric vibrations in one plane. passes through an
optically active substance. the light emerges with its transverse electric vibrations in a
different plane; the cross sectional pattern of the plane polarized light having effectively
been rotated through a definite angle. The angle of rotation can be measured using a
polarimeter, a device which simply has another polarizing material (analyzer) in the path
of the emerging rotated beam. By measurïng the rotation of the analyzer which is required
to confonn to the new plane of polarized light, a measure of the rotatory power of the
sample is obtained. A clockwise rotation of the initial plane polarized light indicates the
sample is dextrorotatory; likewise, a counter-clockwise rotation implies that it is
levorotatory. Since it is the interaction of the light with the individual optically active
molecules which causes the rotation. the rotatory power depends on the number of
molecules in the sample. Consequently, the measure of optical activity which is
conventionally used. is the specifie optical rotation angle, [a] according to the following
equation:
[ ]20 a
aO=lxd (1.2)
where, 1is the path length of the sample cell, and d is the density of the pure liquid or the
concentration of a sample solution, for a sample measured at 20 oC (in this case) using a
1-11
•
•
Generallntrodllction
polarimeter with a monochromatic polarized sodium D-line (wavelength =589 nm) light
source.
Most substances do not effect a net rotation of plane polarized light hecause for
each orientation of a molecule in the sample. there exists the identical. superimposahle
mirror image of that molecule canceling any polarizing effect of the former. Molecules
which are not superimposable with their miITOr images are cal1ed enantiomers. or
stereoisomers. The two optically active forms of the molecule are thus different in the
way in which they behave in a chiral medium. e.g. in plane polarized light, where each
effects an equal and opposite rotation of the beam. Unles.~ an equal number of opposite
enantiomers exists in the sample. there wil1 be a net rotation of the plane polarized light.
A molecule possessing a carbon atom bound to four different ligand~. Le.• a chiml carhon.
presents a requisite element of asymmetry for optical activity. Other tetravalent atoms
may also serve as chiral centers for the enantiomeric configurations of ligands. Sorne
molecules possess more than one chiral center and may. as a result of the canceling-out
effects of the polarization. exhibit no overall optical activity. Le.• they arc achiral. In
addition, there are sorne long chain polymers which do not possess any chiral centers but
display optical activity by virtue of their stiff conformation.
1.5.3 Stereoisomerism Nomenclature
It is necessary to be able to ascertain the absolute configuration of ligands on :i
chiral center as weil as to specify the direction of rotation of the plane polarized Iight
which it effects. The configuration of a molecule is defmed by its chemical structure; it
can be altered only by breaking and reforming chemical bonds. Il is important to note that
while the absolute configuration can be readily ascertained by applying a set of simple
rules, the rotatory power and sign can only he determined experimentally.4
One isotactic polyenantiomer of each of the polyethers used in the sl'.ldies
presented in this thesis are presented in Figure 1.3. The absolute configuration (either R
1·12
General Introduction
or S) of the chiral carbon can he assigned following the sequence mIes put forth by
Cahn et aPi as follows: The chiral carbon ligands are given a priority which depends on
the atomic number of the substituent atom. The molecule is oriented so that the atom with
the least priority (H in both cases in Figure 1.3) is facing away from the observer. A path
is traced, connecting the rerilaining ligands on the page, in the direction of highest to
lowest priority. If the direction of the trace is c1ockwise, the molecule is assigned an R
(ReclUs) configuration; likewise, for a counter-clockwise direction, the ahsolute
configuration is S (Sinister).
R
CH2CtH H H CH2Cl H H H\.l \/ \/ \ ..
C 0 C C 0 C·
cl "C/ "C/ "0/ " / "C/ ",.\. 1\. / ...... 1\ 0H H H ëH2Cl H H H CHP
poly(epichlorohydrin)
s
poly(propylene oxide)
Figure 1.3 Isolactic chain configuration of poly(epichlorohydrin) and poly(propylene
oxide) demonstrating the absolute configurations of the chiral carbons on
each chain.
1-13
•
•
•
General 1nrroducrion
The polymers in Figures 1.3(a) and (b) are stereoregular polymers and are labeled
poly(R-epichlorohydrin) and poly(S-propylene oxide), respcctively, based on the absolute
configuration of the chiral carbon. If the rotation angle of the pure polyenantiomer in
solution has been detennined experimentally, the molecule may be classified further.
Accordingly, the sign of the rotation is included as a prefix to the molecular label: (+) or
d, but not both, is used when the plane of polarized light is rotated to the right; (-) or 1 is
used when the plane of polarized light is rotated to the left The symbols d and 1refer to
the dexrrorotatory and levorotatory rotation directions. respectively. It is important to
remember that the absolute configurations of Rectus and Sinister are unrelated to the sign
of the optical rotation angle and, whereas the absolute configurations never change
without changing the identity of the molecule, the sign of the rotation may vary depending
on the experimental conditions of measurement
1.5.4 Optically Active Polymers
Essentially, there are two types of optically active polymers:2S (i) Main-chain
chiral polymers are optically active due to either the configuration of chiral centers along
the backbone of the chain, or simply to the conformation of the chain which does not
contain any chiral centers. (ü) Side-chain optically active polymers contain chiral centers
in the pendent groups. Traditionally, chiral polymers have been used in the study of the
mechanisms of polymerization reactions,22 and in the investigations of chromatographic
separations,23 and conformation analysis in solution.24 As advances have been made in the
asymmetric synthesis and catalysiS,2S,26 and in enantiomeric resolutions, optically active
monomers and polymers are now becoming increasingly available for their use in other,
recent areas of intrigue. Namely, optical activity has been recognized as a possible
analytical tool to relate the structure of chiral macromolecules to their chiroptical
propertiesP·29 The best example of the relationship between chiroptical properties and
ordered secondary structures is in nature. AlI living species are composed of amino acid
1-14
General Introduction
building blocks which are optically active moleclÙes. Moreover. il is a remarkable làct
that ooly one enantiomer of the amino acids is found in nature. The indication that a
natural selection process has evolved. initiated by chance or otherwise. can ooly imply that
the propeny of optical activity of a substance places restrictions on its hehavior and
interactions with other moleclÙes. To this end. the inherent hehavioral and interaction
restrictions of a synthetic optically active polyenantiomer can he regarded as a useflÙ tool
in the study of the factors goveming its fusion, crysta1lization. and solid state morphology.
1.5.4.1 Poly(epichlorohydrin)
The lowest energy conformation of each polyenantiomer of PECHlo contained in
Figure 1.4 provides a typical example of how the main-ehain chirality of a polymer can
influence the conformation, such as a one-handed helical conformation of the chain upon
condensation. In Figure 1.4(a), the lowest energy conformation of the poly(R
epichlorohydrin) (PRECH) is shown as a helix with a definite sense. Staning at the
bottom of the moleclÙe and traveling up the helix, while tracing a helical line connecting
the adjacent (green) chlorine atoms, leads one in a counter-elockwise sense. In Figure
1.4(c). [(b) is more appropriately treated last] the lowest energy conformation of poly(S
epichlorohydrin) (PSECH) is given as a clockwise sense helix. Again, the sense of the
helix is most easily ascenained by tracing the helicalline which traces the chlorine atoms.
In this case, it is necessary to move from the fust chlorine in the back of the chain in a
winding motion around the back of the chain and up towards the second chlorine atom in
a clockwise motion. The helices in Figures 1.4 (a) and (c) can he termed a left-handed up
R-helix, and a right-handed up S-helix, respectively. The helix in Figure 1.4(b) is included
to answer the question which arises from such assignments: Is a left-handed down R-helix
equivalent to a right-handed up S-helix. It is determined from a close inspection of the
helices in Figure 1.4(a) and (b) that the answer is no. The solution is most easily reached
ü the helices are inspected by tuming the page on its side. The relative direction of the
1-15
•
Figure 1.4
Genera/Introduction
~ ·~l ·l. .J' ~'.
'.·
'~f .' .~ .~l ' ..
· l ~. '~r
~ .~l .1". .'.
• J ). · '~r, .~ .~l ·l..
'.· J ) · '~r~, .
-1:.~ .~l •••
· J. ) · '~r
A B C
Computer modeled lowest energy conformation of (a) an upward poly(R
epichlorohydrin) helix, (b) a downward poly(R-epichlorohydrin) helix, and
(c) an upward poly(S-epichiorohydrin) helix.
1-16
•
•
•
General Introduction
chlorine atoms is immediately evident and it is found that, not only is the direction of the
overall dipole moment (detennined by the pointing direction of the chlorine atoms) in
opposite directions for the helices in question, but, remarkably, the direction of the overall
dipole moment in the R-helix (up) and the S-helix (up) are the same.
1.6 The Present Work
The discovery of new and intriguing struclUre-property relationships among
optically active polymers constilUtes much more than an effort to meet our changing
needs. Indeed, the slUdies involving the unique applications of macromolecular chirality
will continue to refine the fundamental understanding of macromolecular materials in
general and stimulate further slUdy.
It is therefore the aim of this thesis to employ the property of backbone
macromolecular chirality to gain a deeper understanding of the complex processes
associated with the fusion of polymer molecules; the factors which influence and which
govern the crystallization of polymers; the intricate relationship between the different
order structures in the morphological hierarchy of the final crystalline solid state.
The thesis is divided inta six ehapters:
Chapter One is a general introduction. Each subsequent ehapter begins with a
thorough introduction to the topics speeifieally addressed in that ehapter.
Chapter Two briefly reviews the topie of ring-opening polymerization and eontains
the full experirnental details of the syntheses of bath enantiomers and the stereoblock fOrtO
of optieally aetive poly(epiehlorohydrin) and the subsequent dechlorination of these
1-17
•
•
•
Genera/lnrroduction
polyrners to the corresponding optically active poly(propylene oxide). The results of the
basic characterizations of al! of the polyrners used in the slUdies presented in subsequent
portions of the thesis are also included.
In Chapter Three, an extensive investigation of the thennal behavior of the
polyrners is presented. The results of a multiple melting behavior analysis, using
differential scanning calorimetry, are critical!y discussed in terms of current theory and a
mechanism is proposed for the observed polyrner melting behavior.
In Chapter Four, the topic of crysta1lization in enantiomeric mixtures of optically
active polyrners is reviewed. The results of the investigation of the isothenna1 kinetics of
the growth of melt-crystallized spherulites of the polyenantiomers, blends of enantiomers
and the stereoblock form, obtained by polarized light optical microscopy, are presented.
The data are analyzed in terms of the currently available theory and are related to polymer
structure. The main components of this Chapter have already been published in
Macromolecules.31
In Chapter Five, the subject of banding in polyrner spherulites is discussed in light
of the results of the detailed investigation of the spherulite morphology using mainly
atomic force microscopy. This Chapter has also recently appeared in Macromolecules as a
full paper.32
Chapter Six contains a summary of the conclusions which can be drawn from the
work presented in the thesis. The significant contributions made to original research are
noted and several ideas for the proposed continuation of the project are also included.
1-18
(6)
(7)
• (8)
(9)
(10)
(1l)
(12)
(13)
•General Introduction
1.7 References
(1) Young. R. J.; Lovell. P. A. Introduction co Polymers; ehapman & Hall: London.
1991. p.243.
(2) Vaughan. A. S.; Bassett. D. e. In Comprehensive Polymer Seience Vol.2; Booth,
e.; Priee, e. Ed.'s; Pergamon: Oxford. 1989, pAIS.
(3) Natta. G.; Farina, M. Stereochemisrry; Longman: London. 1972.
(4) Morrison. R. T.; Boyd. R. N. Organic Chemisrry; Allyn and Bacon, Ine.: Boston.
1983.
(5) Mandelkem, L. An Introduction co Macromolecules; Springer-Verlag: New York.
1964.
Keller, A. Philos. Mag. 1957,2,1171.
Till. P. H., Jr. J. Polym. Sei. 1957.24,301.
Fischer, E. W. Z. Naturforsch.• Teil A 1957, 12, 753.
Mandelkem, L. J. Phys. Chem 1971, 75. 3920.
Wunderlieh, B. Thermal Analysis; Academie: London. 1990.
for example. Bisault, J.; Rysehenkow, G. J. Crystal Growth 1991. 110, 889.
for example, George, J.; Premaehandran. S. K. J. Crystal Growth 1979. 46, 297.
Priee, F. P. In Nucleation; A. e. Zettlemoyer, Ed.; Marcel Decker: New York,
1969.
(14) Norton, D. R.; Keller. A. Polymer 1985, 26, 704.
(15) Geil, P. H. PolymerSingle Crystals; John Wiley & Sons: New York, 1963.
(16) Hoffman, J. D.; Davis, G. T.; Lauritzen, Jr., 1. J. In Treatise on SoUd State
Chemisrry; Hannay, N. B., Ed.; Vol. 3; Plenum: New York, 1976; ehapter 7.
(17) Keller, A. J. Polym Sei. 1959.39,151.
(18) Fujiwara. Y. J. Appl. Polym Sei. 1960,4,10.
(19) Keith. H. D.; Padden, F. J. Jr.; Russell, T. P. Macromolecules 1989, 22, 666.
1-19
•
•
GenerallnrrodllcTiclII
(20) Shurcliff, W. A.; Ballard, S. S. Polarized LighT; Van Nostrand: Princeton, 1964.
(21) Cahn, R. S.; Ingold, C. K.; Prelog, V. Angel\'. Chem. Imern. Ed. Engl. 1966. 5,
385.
(22) Selegny, E., Ed. 0pTical!y ACTive Polymers; Reidel: Dordrccht, Thc Nctherlands,
1979.
(23) A11enmark, S. G. Chromatographic EnamioseparaTion; Ellis Horwood Ltd.: Ncw
York,1991.
(24) Lifson, S.; Andreola, C.; Peterson, N. C.; Green, M. M. J. Am. Chem. Soc. 1989,
111,8850.
(25) Belfie1d, K. D.; Beltield, J. S. Trends Polym. Sei. 1995, 3, 180.
(26) Moore, 1. S.; Stupp, S. 1. J. Am. Chem. Soc. 1992, 114, 3429.
(27) Wulff, G. Angell'. Chem.. lm. Ed. Engl. 1989, 28, 21.
(28) Williams, D. J. Angell'. Chem.. lm. Ed. Engl. 1984,23, 690.
(29) Green, M. M.. Peterson, N. C.; Sato, T.; Teramoto, A.; Cook, R.; Lil:~on, S.
Scienet 1995, 268, 1860.
(30) Pcrsona) Communication: Ilie Saracovan, McGiU University, 1995.
(3 I) Singlicld, K. L.; Brown, G. R. Macromolecules 1995, 28, 1290.
(32) Singlield, K. L.; Klass, J. M. K.; Brown, G. R. Macromolecules 1995, 28, 8006.
1-20
POLYMER SYNTHESIS& CHARACTERIZATION
Chapter Two
2.1 Introduction
Ring· Opening Polymerization
2.1.1 General
The polymerization of cyclic ethers to linear polymers is arnong the most versatile
polymerizations known. Ring-opening mechanisms can yield heteroatom main chains of
high molecular weight typically associated with addition polymerization mechanisms of
bifunctional monomers.1 Polymerization reactions, in general, are exoentropic processes
since they involve the ordering of individual monomers into long chains. This unfavorable
change in entropy must he accompanied by a decrease in enthalpy for the polymerization
to occur. In ring-opening polymerizations, the negative enthalpy term arises t'rom the
release of the ring-strain energy of the monomer. For three-memhered heterocycles. this
energy is comparable to that available from double bond opening, as in vinyl monomers.
Ring-opening polymerizations are classified according to type of reaction mechanism.
Oxiranes are three-memhered cyclic ethers (epoxides) which, along with their
derivalives, exhibit high kinetic reactivity and consequently are polymerizable by both
anionic (nucleophilic) or calionic (electrophilic) mechanisms. By contrast, larger rings are
2-1
•
•
Synrhesis & Characreri~arioll
usually only reactive to the cationic mechanism; the initiators are represented mainly hy
alkali metal compounds, and will not he discussed. The terms initiator and c:lta\yst are
used interchangeahly in this chapter.
Polymerizations of oxiranes with anionic initiators proceed via metal alkoxides.2.•
The most significant feature of the anionic process is a possihle coordination inter.lction
between the oxygen of the monomer and the metal atom of the initiator. A coordinate
interaction is favored when the metal atom of the initiator is of strong Lewis acid
character. To this end. the anionic initiators can he further classilied into two groups:
simple anionic and coordinate: (1) Simple anionic initiators are gener.llly quite effective
for the polymerization of non-substituted oxiranes. while attempts to polymerize
substituted ones are in general futile due to the occurrence of side reactions.
Monosubstituted oxiranes are usually polymerized without stereo control ta soluble,
amorphous products.1 Simple anionic initiators will not he discussed further.
(2) Coordinate initiators are characterized by the high Lewis acidity of the metal atom.
Metal atoms of the coordinate-type initiators are typically Group Il or III metals, such as
zinc or alurninurn and iron is expected to hehave in a coordinate làshion as weil. The
coordinate catalysts mentioned in this introduction initiate polymerization of the type
where the growing chain end is presumably negatively charged and thus are initially
c1assified in this anionic group. However, to avoid confusion. the qualiliers cationic and
anionic are avoided, since there are both cationic and anionic sites present during
coordination polymerization. Coordination catalysts offer the potential to polymerize
monosubstituted oxirane monomers to bigh molecular weight, crystalline. stereoregular
polymers.
The numher and diversity of ring-opening polymerization reactions is reflected in
the enormous amount of Iiterature in this field.s It is therefore far beyond the scope of this
introduction to review the subject in detail. Instead. a brief and selective overview with a
2-2
Synthesis & Characrerizarion
historical perspective is presented, with the focus on the coordination catalyst selected for
the polymeri7.ation of the optically active epichlorohydrin monomers used in this thesis.
2.1.2 Stereoregular Polyethers
To begin, the concepts of stereo- and regio-control during ring-opening
polymerization are addressed. Unlike vinyl monomers, where chiral centers are generated
during the polymerization process, methyl-substituted oxiranes, (b) and (c) below, are
optically active monomers which can exist in two enantiomeric forms. After
polymerization, the asymmetric centers (marked below with a *) exist as true chiral
centers on the polymer backbone. The configuration of the chiral carbon of the monomer
is sensitive to the ring-opening attack at that carbon. In principle, a racemic monomer
mixture can be polymerized with control of stereoregularitiy either by a stereoselecrive or
a stereoelecrive type of mechanism:6•7
oxirane(ethylene oxide)
(a)
~/o,~H2C-CH
1CH3
methyl oxirane(propylene oxide)
(b)
o/ '*H2C-CH
1CH2CI
chloromethyl oxiran(epichlorohydrin)
(c)
(l) A stereoselective mechanism is one in which each enantiomer of the racemic monomer
is selectively incorporated into its own type of polymer chain such that polymerization of a
racemic monomer yields an isotactic, statistical mixture of R-chains and S-chains. In this
case, the enantiomeric composition of the polymer is at ail limes identical to that of the
monomer. (2) ln a stereoelective polymerization, the catalyst specifically incorporates one
enantiomer of a racemic monomer mixture into the growing polymer chain. Operation of
2-3
•
•
•
Synthesis & Characterization
this mechanism leaves the unincorporated enantiomcr monomcr hchind. unrcacted.
Stereoelective catalysl~ are at least to sorne extcnt chiral. and many are prepared by the
use of chiral ligands.
In practice. it is rare for any particular catalyst system to show complete
stereoselection or stercoclcction. and most catalysl~ give complex hchavior which i~
dependent not only on the type of catalyst but on the rcaction conditions.6 Typically.
these stereo-control\ed polymerizations will produce stereoregular polyrners with few
defect sites which covalently link long stereo-sequences of opposite conliguration.
Due to their a~ymmetry. the substituted oxirane monomers can he incorponlled
into the growing polyrner chain either by a regioregular head-to-tai! addition. or by an
irregular head-to-head or tail-to-lail addition. To produce stereorcgular polyethers l'rom
the methyl-substituted oxiranes. ring-opening at the ~-carbon is rcquired, exclusivcly.
Ring-opening at the ~-carbon leaves the chiral center intact and ultimately mainlains the
interrelation of monomer and polyrner enantiomeric composition. In this respect, a
random site attack will produce a regioirregular polyrner which will also he stereorandom.
Thus, the combination of stereo- and regio-control during polymerization is required to
produce an isotactic. crystalline polyether.
2.1.3 Historical Perspective
2.1.3.1 Coordination Catalysts
Oxirane polymerization is one of the oldest examples of synthetic macromolecules
production. In 1863, Wurtz was the first to report the formation of oxirane oligomers
based on ethylene oxide.8 The polyrnerization of methyl oxirane was frrst reported by
Levene9 in 1927, and the frrst high molecular weight polyether, poly(ethylene oxide)
(PEO), was reported by StaudingerlO in 1933. In 1955, Nattall reported that the use of
Ziegler-type catalysts. consisting of organometallics such as a\kylaluminums, combined
2-4
Symhesis & Characterization
with transition metal compounds such as TiCl., resulted in a crystalline polypropylene.
This polymer was described as having successive pseudo-asymmetric centers of the same
configuration, i.e., stereoregular, and the phrase isocactic was coined. 12 In the case of
these Ziegler catalysts, the yield of crystalline polypropylene was low relative to the
amorphous fraction. l ) In 1955, Vandenberg and his colleagues discovered that under
certain conditions, reaction of propylene in the presence of the TiCl) solids with large
amounts of AlCI) solids, presumably mixed crystaIs, gave high yields of the crystalline
polypropylene.l • Also in 1955, Pruit and Baggett1S reported the polymerization of the
asymmetric monomer, propylene oxide, by a new catalyst consisting of the reaction
product of FeCI) and propylene oxide monomer. This catalyst polymerized the propylene
oxide to a new crystalIine polymer of a high molecular weight along with an amorphous
rubber fraction. The significance of this discovery is reflected in the fact that this catalyst
is routinely used today.
Il was Price and Osgan who proposed the now well-accepted mechanism of
coordination polymerization for the Ziegler-Nana polymerizations of propylene.16
Moreover, they suggested that coordinate-type interactions occurred during the Pruitt and
Bagget polymerization. Price and Osgan also performed the flfst mechanism studies
specilically on the stereochemistry of oxirane polymerization in 1956.16 The
polymerization of optically active propylene oxide was chosen as the subject of
investigation since. of the monomers already sludied. propylene oxide was unique in
possessing an asymmetric center both before and after polymerization. The producls of
polymerization of the Pruitt-Baggett FeCl)-propylene oxide initiator were compared with
those produced with the ordinary base catalyst. potassium hydroxide. The results of their
investigation are summarized in scheme 1.
Polymerization of the optically active monomer using KOH yielded an optically
active. isotactic crystalIine polymer of low molecular weight. However. with the Pruitt
and Baggett catalysl. polymerization of the optically active propylene oxide gave two
2-5
Synthesis & Characterization
different poly(propylene oxide} (POO) fractions of similar, high mo1<:cular weighl. One of
the fractions was a crystalline, optically active polymer which gave the same X-ray pallern
and the same specifie optical rotation angle as tht ';ptically active crystalline polymer
obtained by the use of KOH initiator. The indistinguishahle X-ray patterns demonslralcd
Scheme 1
Crystalline polymer[a] = +25°, CHC\)
Crystalline polymer[a] =+25°, CHC!)
Amorphous polymer[a] = +3°, CHC\)
that the crystalline product obtained using Feel3 was isotactic and that its asymmetric
carbon centers retained the configuration of the monomer from which it was derivcd. The
second fraction ;vas an amorphous polymer, and on the hasis of its rotation angle. a
stereor:mdorn !,ul)'!!ler. The apparent anomaly of the generation of racemized polymcr
from optically active monomer was later explained, by examination of the diol fragmenl~
of the cleaved polymers.4 It was shown unequivocally that the minor amorphous fraction
obtained from propylene oxide polymerization with a Pruitt and Baggett catalyst contained
substantial head-to-head and tail-to-tail regioirregularities along with the expected head
to-tail sequences.
Similar experiments were performed on the polymers obtained l'rom
polymerization of methyl-disubstituted oxiranes which contain two chiral centers per
monomer.4 Examination of the cleavage products of these polymers made it clear that
there was an inversion of configuration at the asymmetric center undergoing riilg-opening.
2-6
Synthesis & Characterization
The experiments also demonstrated that propagation is possible with a coordination
catalyst by ring-opening at the lX-carbon of a monosubstituted oxirane.4 Polymerization of
an optically active propylene oxide monomer by ring-opening at the lX-carbon gives
inversion of configuration and thus racemization is observed in the polymer. Apparently a
sufficiently sterically hindered site in a coordination catalyst assembly is required to yield
solely head-to-tail polymerization (Le., propagation solely by ~-carbon attack) of
propylene oxide.4
2.1.3.2 Alkyl Aluminoxane C8talysts
Price and Osgan also developed a coordinate-type aluminum isopropoxide-ZnClz
catalystP which polymerized propylene oxide monomer to a crystalline. high molecular
weight polymer. Vandenberg and bis colleagues speculated on the potential catalytic
similarity between the new Price and Osgan catalyst and sorne of their own newly
developed transition metal-based vinyl ether initiators.18 With die aim of generating a new
type of crystalline polymer, they attempted the polymerization of the epichlorohydrin,
using their own initiators. A small "mount of the new type of crystalline polyether,
poly(epichlorohydrin), was obtained.19
There is an interesting anecdote which accompanies the development of these
original initiators to the alkyl aluminoxane catalysts used today:zo In the course of
optimizing the conditions, Vandenberg improved the yield of poly(epichlorohydrin) with a
catalyst consisting of triisobutyl aluminum and vanadium trichloride; however, the major
product was an amorphous rubber. Although this product would eventually emerge as an
elastomer of industrial importance, they had obviously encountered reproducibility
problems with the new bateh of monomer. Initially. a nearly empty bottle of
epichlorohydrin monomer had been used, and it was thought that the contents might have
been contaminated with water. To test this hypothesis, triisobutyl aluminum was reacted
with sorne water, and the product was found to be very effective catalytically for the
2-7
Synthes;s & Character;zat;o1/
polymerization of epichlorohydrin monomer to the crystalline polymer. ln this fonuitous
manner, Vandenberg discovered an important class of new alkvl alllm;noxane cawl\'.I't.l'. .
based on the reaction of alkyl aluminum with water. ln an allcmpt to prove the
coordinating feature of the catalyst during polymerization, a ehelating agent, chosen as
acetylacetone, was added to block the coordination site on the aluminum and cl1cctively
inhibit polymerization. In fact, the chelating agent improved the pcrformancL and the
modified catalyst was referred to as a chelating catalyst4 A schematic diagram of the
Vandenberg alkyl aluminoxane catalyst formation is shown in the following scheme 2.4
Scheme2
Others have systematically investigated the role of water in these coordination catalysL~.
Furukawa et al.21 reported that the soUd surface-active diethylzinc catalyst is actually
polynuclear and has more than one Zn-O-Zn grouping for polymerizing propylene oxiœ to
high molecular weight polymer. Colclough et al.22 found that trimethylalumin·.1m must he
reacted with water to give an active catalyst for propylene oxide polymerization,
particularly for the preparation of a crystalline PPrO fraction.
2·8
•Synthesis & Characrerizarion
Vandenberg continue<! with the development of other coordinate catalysts l'rom
simple organometallic compounds and found that by reaction with a/moS( stoichiometric
amounts of water, alkyls of a1uminum, zinc, and magnesium can be activated for
polymerizing oxiranes.23 The optimum amount of water was about one mole per mole of
metal atom for the magnesium and zinc a1kyls, and 0.5 mole per mole of metal atom for
the a1uminum a1kyls. The a1uminum catalysts are unique because they are stable if kept
free l'rom air and moisture, and particularly because they are soluble in organic solvents.
By contrast, the magnesium and zinc catalysts, like the Ziegler-type catalysts, are
heterogeneous.
2.1.4 Mechanism of Alkyl Aluminoxane Catalyst Polymerization
In oxirane polymerization, the alkyl a1uminoxane catalyst fol1ows a coordinate
anionic propagation mechanism similar to that suggested previously for sorne epoxide
catalysts as weil as Ziegler catalysts.16 The proposed initiation and propagation pathways
for the polymerization of a monosubstituted oxirane are ilIustrated in Scheme 3.4
Polymerization is initiated by a reaction between the oxygen on the monomer and the
a1uminum of the alkyl a1uminum bonds to give an Al-O-C-C(R)-alkyl grouping. The
monomer is coordinated to one of the a1uminum atoms prior to its attack by the growing
chain on the other a1uminum atom, so that each propagation step is preceded by a metal
epoxide coordination step.
Scheme3
R C-R1 1 \
-C-C-(}·,c-d)1:1 ~
";~<Al
1\
..,\-C-Rd
R R1 1
R-C ç-C~-C-C1--
-1;f\:?";~~AI(
1Al
1\
2-9
Synrhesis & Characteri~ati()n
The coordination bonds in Ù1e catalyst structure are Ù1erefore required to shuttle Ù1e
growing polyrner chain from one Metal atom to an adjacent one wiÙ10ut altering the
valence of the Metal. Short of Ù1e in situ elucidation of Ù1e structure of the initiator
assembly, it can only be postulated that the role of the chelating agent in the chelate
catalyst is to minimize ordinary acid or cationic polymerization of some monomers hy
blocking Ù1e fourth coordination position of aluminum and allowing Ù1e IifÙ1 or SiXÙ1
coordination positions to function in the propagation step.4 Altematively, Ù1e same net
effect of decreasing the acidity of the aluminum May be obtained if Ù1e coordination bond
of the chelate is displaced temporarily by the oxirane during each propagation step of Ù1e
polymerization.4
The Rl Al·H20 alkyl aluminoxane catalysts can behave as superior cationic
catalysts with epoxides which can be polyrnerized readily cationically, or as coordination
catalysts with other epoxides such as epichlorohydrin.4 The use of a modilied chelate
catalyst always induces a coordination polymerization mechanism. Epichlorohydrin
polyrnerizes best with the alkyl aluminoxane catalysts because of the reactive chlorine
which apparently leads to catalyst-destroying side reactions with Ù1e more simple basic
catalysts.27 Stereoregularity of the pctlyrner is controlled by steric hindrance at Ù1e Metal
sites, a so-called catalyst site control.4 The growing site can readily change in steric
hindrance to give either more or less stereoregularity. For example, in the polyrnerization
of phenyl glycidyl ether using the chelated catalyst nearly quantitative yieids of isotactic
polyrner were formed.4 This is presumably a case of catalyst site control and it is also an
example of the catalyst operating largely by a coordination rather than a cationic
mechanism. Compared to the relatively small amounts of isotactic polyrner resulting from
the polyrnerization of propylene oxide (20 %), the greater stereoregularity with phenyl
glycidyl ether may reflect the greater steric hindrance by the bulkier side chain and/or it
may be due to the coordination of the ether side chain with Ù1e catalyst to provide a more
hindered site.24
2·10
•Synrhesis & Characterization
ln general, the polymerizations of mono- and disubstituted epoxides with all
known anionic. cationic. or coordinate catalysts occur with an inversion of configuration
of the ring-opening carbon.13,24 It is important to note that in the case of polymerization
of epichlorohydrin, the carbon undergoing the ring-opening is not the asymmetric carbon;
therefore configuration of the chiral center is unaltered. On the other hand, in the case of
coordinate polymerization of propylene oxide, the less hindered ~-carbon can undergo
ring-opening, thus resulting in the formation of a fraction of stereorandom polymer. The
performance of the coordination catalyst varies a great deal with catalyst preparation, the
steric and electronic structure of the monomer, and the polymerization conditions.
The alkyl aluminoxane catalysts were found to he very effective for the
polymerization of many epoxides to crystalline, stereoregular polymers of high molecular
weight. e.g. poly(epichlorohydrin),26 poly(styrene oxide),23 poly(butadiene monoxide),23
poly(epibromohydrin),23 and poly(glycidyl ether).!3 More recently, these alkyl
aluminoxane catalysts were found to also he effective for the ring-opening polymerization
of cyclic lactones, such as ~-butyrolactone. 27
2.1.5 Chemical Modification of Poly{epichlorohydrin)
Steller28 has reported that the high molecular weight crystaIline
poly(epichlorohydrin) produced by the Vandenherg method can he quantitatively
converted to crystaIline poly(propylene oxide) by the replacement of chlorine with
hydrogen using lithium aluminum hydride (LiAlH4) reducing agent, in tetrahydrofuran
(nIF) at 50 oC.
CH,1
-fCH-CH,-O+-* .
The structure of the polymer is not otherwise altered, so the reaction occurs with retention
of configuration of the asymmetric centers. Figure 2.1 is a recollstruction of the reaction
2·11
60 -
•Synthes;s & Characterizat;on
100 -c~}-..---------------.à.,
80 -'0.
"'0
40 -"0
20 -
b
10 +-"T"""-r-I-""--r-I-""--r-I~--r-I~''''''-~
o 4 8 12
Time (h)
16 20
Figure 2.1 Semi-Iog plot showing dechlorination of poly(R,S-epichlorohydrin)
elastomer with LiAlH4 (2.5 moleJmole Cl) in TIIF at 50 oC (taken from
reference 28).
2-12
Synrhesis & Characterization
rate data for the dechlorination of poly(R,S-epichlorohydrin) as reponed by Steller.28
Mter a reaction lime of 50 h, elemental analysis of the resulting polymer gave no sign of
any remaining chlorine. It is interesting to note that a1though the asymmetric carbon
remains unchanged, the modification of tht: substituent group effectively reverses the
absolute configuration of the chiral carbon. That is to say, in the process of dechlorination
the poly(R-epichlorohydrin) (PRECH) is convened to poly(S-propylene mode) (PSPrO).
Likewise, the poly(S-epichlorohydrin) (PSECH) polymer is convened to poly{R-propylene
oxide) (PRPrO).
The asymmetric epichlorohydrin monomer is commercially available in both pure
enantiomeric forms. Each optically pure epichlorohydrin monomer is readily polymerized
with the coordinate Vandenberg catalyst [(C2Hs)3Al-O.6H20j to produce each optically
pure, regioregular polyenantiomer. Subsequent dechlorination of each polyenantiomer
should give the corresponding optically pure, regioregular polyenantiomer of
poly{propylene oxide).
2.1.6 The Present Work
In the following sections of this chapter, the experimental procedures are outlined
for the polymerization of optically pure and raccmic epichlorohydrin monomer using the
triethylaluminum-water catalyst. The procedure for the dechlorination of
poly{epichlorohydrin) (PECH) to poly{propylene oxide) (PPrO) is also presented. The
Experimental sec~on of this thesis bas been written with attention to detail. Althougb the
methods herein are not original, the purpose of the supplementary details and comments is
to provide a practical, complete reference for any future polymerization and fractionation
methods used with this versatile initiator. The results of the basic polymer
characterization experiments of the newly synthesized optically active polymers, their
blends, and the polymer from the polymerization of a racemic monomer are described.
2-13
•Synrhesis & Characterization
2.2 Experimental
2.2.1 Materials
Racemic epichlorohydrin monomer (99+ %, Aldrich) was distilled heforc use,
discarding the first 100 mL. The optically active monomers (R)-(-)-epichlorohydrin (99 %
Aldrich, [«lio =_34°) and (S)-(+)-epichlorohydrin (99 %, Aldrich, [«lio =+34°) were
used as-received. A 1.0 M solution of triethylaluminum in hexane (Aldrich) was used in
the preparation of the polymerization catalyst. Only freshly opened, anhydrous diethyl
ether (BDH) was used in all reactions. In the dechlorination of the poly(epichlorohydrin)
to poly(propylene oxide), lithium aluminum hydride, LiAlH4, was used either in pellet
form (Aldrich) or as a 1.0 M solution in tetrahydrofuran (Aldrich). The THF (BDH) used
in the dechlorination reaction was distilled over sodium henzophenone complex, collected
under nitrogen, and used immediately.
2.2.2 Methods: General Overview
A general overview is presented of the procedures used in the synthesis of
poly(epichlorohydrin) and the subsequent conversion of this polymer to poly(propylene
oxide), which are described in detail in the following Sections 2.2.3-8.
• Sajety AlI of the procedures were carried out under an inen atmosphere, either on
a vacuum line or in a glove bag, and with the use of flame-dried glassware. Extreme
caution was exercised in the handling of the very reactive Et3Al solutions, and the
epichlorohydrin monomer wlùch is a known mutagen and a cancer suspect agent.
• Catalyst Anhydrous diethyl ether was used to dilute the as-received solution of in
hexane by a factor of IWO. The resulting Et3Al solution was brought to a reflux and the
catalyst was activated by adding the almost stoiclùometric amount of water dropwise,
during the reflux. The final activated catalyst solution was clear and stable at room
temperature, and stored under an inert atmosphere.
2-14
•Synthesis & Characterization
• Polymerizotion Polymerization of the epichlorohydrin monomer using the above
catalyst solution was camed out in anhydrous diethyl ether at 25-30 oC over a period of
ca. 15 h. The reaction was stopped by the addition of anhydrous ethanol, and the
insoluble polymer was filtered out of the product slurry. The residuai cataiyst was
removed by al\uwing a methanol/hydrochloride slurry of the polymer to stand for the
appropriate rime period. The solid polymer was repeatedly washed and filtered until
neutral.
• Fractionation The acetone soluble atactic materiai was removed from the crude
•
•
poly(R,S-epichlorohydrin) producl. The remaining acetone insoluble fraction was refluxed
in acetone and allowed to cool slowly to room temperamre. The precipitated isotactic
fraction was collected. No fractionation of the product of the polymerization of the
opticaily pure epichlorohydrin monomers was performed.
• Dechlorination The dechlorination of poly(epichlorohydrin) to poly(propylene
oxide) was performed with the reducing agent,~, in THF solution at 50 oC for 50
12 h. The reaction was stopped by the addition of anhydrous ethanol at low temperamre
(ca. 0 oC). Solid residues were removed by filtering the polymer solution, which was then
evaporated to dryness, and the polymer was redissolved in benzene and freeze-dried.
• Storage AlI of the polymers used in the smdies described in this thesis were stored
in vacuo and no antioxidants were added.
2.2.3 Formation of EtsAI·O.6H20 catalyst
2.2.3.1 Apparatus
The schematic diagram in Figure 2.2 shows the assembly of required glassware
attached to a vacuum line. A 2 L, 5-neck round-bottom reaction flask was equipped with
a condenser, a mechanicai stirrer (Bantant Series 20 Mixer used with a glass shaft and a
Teflon blade), and a caiibrated addition buret. A thermometer well in the flask was filled
2-15
•Synthesis & Characteri:ation
ta vacunun pump
'lOnIge nul<~ drying
column
10H&l'lI!l'Hli4 diffusion_-"'"'ï pump
•
Hg bubblcr
Figure 2.2 Schematic diagram of the polymerization apparatus.
2-16
•Synthesis & Characterization
with silicone oil and a resistance temperalUre detector (RTD) probe was placed inside the
weIl. The top of the condenser was fitted with a T-type dry nitrogen input. Zero oxygen
research grade nitrogen gas (Matheson) was flfst passed through a 1 m x 2.5 cm column
filled with molecular sieves, of mesh size 4A (Aldrich). A vacuum stopcock was fused to
the fifth neck of the reaction flask and a wired-down rubber septum was placed on the
opening of the barrel. A heating mantle was placed under the reaction flask.
The vacuum stopcock on the fifth neck of the reaction flask was opened and the
septum on the barrel was pierced with a 14 gauge, 5 cm-long syringe needle. The needle
was used to thread the septum with about 60 cm of 20 gauge Teflon flexible needle tubing
(Aldrich) which was terminated with a stainless steel KEL-F luer hub. The tubing was
inserted through the barrel until it reached the stopcock, leaving sufficient space to close
the stopcock. The stopcock was left open. The opposite end of the tubing (a female luer
hub) was equipped with a male-to-male luer hub adapter with mini stopcock (Aldrich).
The mini stopcock was closed. The system was purged repeatedly using dry nitrogen and
vacuum, and then left under a dry nitrogen atmosphere. The vacuum stopcock on the fifth
neck of the reaction flask was opened briefly to purge the barrel and the flexible tubing
with nitrogen.
2.2.3.2 Dilution of E~Alln Hexane
Since trlethylaluminum in hexane reacts violently with air and moisture, it is
imperative to maintain a dry, oxygen-free environrnent and to flame-dry ail of the
glassware and the stainless steel needles prior to use. Inside a nitrogen atmosphere glove
bag, in the fume hood, 400 mL of fresh, anhydrous diethyl ether was measured using a
1000 mL graduated cylinder, and added to a 1000 mL I-neck round-bottom flask (ether
storage flask). A wired-down rubber septum was placed on the f1ask and it was attached
ta the vacuum line. A partial vacuum created in the reaction vessel was used to transfer
2-17
•Synrhesis & Characteri~arion
200 mL ether from the storage l1ask to the reaction l1ask through the septum un the
addition buret using a stainless steel double-tipped needle (Aldrich).
The glove bag in the fume hood was equipped with a dry nitrugen inlet from the
vacuum line via Tel10n tubing terminated by a 5 cm stainless steel syringe needle. This
nitrogen was used for liquid transfers in the glove bag and not for the inl1ation of the glove
bag. The following items were then placed into the glove hag: a clean. dry 500 mL
storage l1ask equipped with two vacuum plugs and stoppered with a wired-down ruhbcr
septum; the 1 L Sure Seal boule of a 1.0 M solution of triethylaluminum in hexane
(Aldrich); polypropylene disposable syringes (20 and 30 mL size); a stainless steel double·
tipped needle; a top-Ioading analytical balance; a buret stand with clamp; a male-to-male
luer adapter. The glove bag was purged three times with nitrogen. The top vacuum plug
of the storage flask was opened. The storage flask was c1amped ooto the huret stand and
the solution bottle was placed on the balance. The septum on the bottle containing the
Et3Al in hexane was pierced with the dry nitrogen inlet needle. Using the density of the
Et3Al solution. and monitoring the change in weight of the solution boule during the
transfer. the required volume (400 mL) was transferred with nitrogen pressure to the
storage flask. using a double-tipped needle.
The storage flask containing the Et3Al solution was added to the assembled
glassware on the vacuum line. The flask was periodically opened to the dry nitrogen
atmosphere of the vacuum line. The rubber septum on top of the addition buret was
removed and replaced with the storage flask containing the Et3Al in hexane. The addition
buret was opened to the vacuum and then flushed with the dry nitrogen. Ali of the Et3A\
solution was added dropwise. over a one hour period. into the reaction flask containing
the diethyl ether. with slow stirring. When a\l of the solution was added. the addition
buret was isolated from the system and it was evacuated. One end of the double-tipped
needle was inserted through the septum of the empty storage flask atop the addition buret.
and the other end pierced the septum of the ether storage flask. The remaining 200 mL of
2-18
•
Synthesis & Characrerizarion
diemyl emer was drawn into me rcaction fiask by the vacuum in me addition huret.
Finally, the reactor was opened ta me addition buret, a:ld the remaining ether solvent was
addcd ta the rcaction fiask wim stirring.
2.2.3.3 E~AI Reaction with Water
The storage flask on top of me addition buret was removed and replaced with a
rubber septum, secured wim copper wire. Approximately 20 mL of twice-distilled,
deionized water was added to the addition buret through the septum using a syringe.
Exactly 4.3 mL of water was added dropwise to me reaction solution, wim no more man
one drop being released from the graduated addition buret at one time. The addition of
water to the reaction fiask eaused a violent evolution of cthane. The nitrogen flow over
me condenser was kept to a nùnimum to avoid me loss of emer through the oil bubbler
with me exit of me emane gas. Since cold tap water did not provide adequate cooling,
commercial antifreeze cooled to 5 oC in a refrigerated 8 L bam was circulated through the
condenser.
After completion of me water addition, me reaction solution was set to reflux,
keeping the nitrogen flow over the condenser to a nùnÎmum. After refluxing for 2 h. the
reaction solution was cooled to room temperature and could he left safely ovemight
before it was transferred to the catalyst storage flask.
2.2.3.4 Storage of the Catalyst Solution
The storage flask with two vacuum stopcocks was used as the catalyst storage
fiask. The vacuum stopcock on the f1fm neck of the reaction flask was opened (wim the
nùni stopcock at the opposite end of the Teflon tubing still c1osed) and me Teflon tubing
was pushed through the septum unill the end was submerged near to the bottom of the
reaction flask. The nùni stopcock at the other end was attached to a 5 cm-long stainless
steel syringe needle. which was used to pierce me septum of the evacuated catalyst
2-19
•
•
•
Synthesis & Characteriuuion
storage flask. The mim stopcock was immediately opencd to pennit the tnUlsfer of ail of
the catalyst solution from the reaction flask to the catalyst storage flask. The fina1
concentration of ,he catalyst solution was approximately 0.5 M Et3Al.
The clear catalyst solution could be safely stored on the vacuum \ine, and remaincd
stable ifperiodically flushed with the dry nitrogen atrnosphere. Signs of catalyst instability
include the formation of so\ids on the sides of the flask, yellowing of the solution, and
pressure build-up inside the flask.
2.2.4 Polymerization of Epichlorohydrin
The glassware was re-assembled as ilIustrated in Figure 2.2. The addition buret
had a rubber septum on top. The glassware assembly was evacuated and purged with dry
nitrogen, repeatedly, and then flame-dried under vacuum with a Bunsen bumer. The
system was left under a dry nitrogen atrnosphere.
In a glove bag in the fume hood, 550 mL of freshly opened anhydrous diethyl ether
was measured with a 1000 mL graduated glass cylinder and added into aiL ether storage
flask. The flask was stoppered with a wired-down rubber septum. Also in the glove bag,
105 mL (125 g) of racemic epichlorohydrin monomer was transferred from the boule to
disposable syringes. On the vacuum line, the reaction flask was evacuated and ail of the
diethyl ether from the storage flask was drawn over, through the septum on the addition
buret using the double-tipped needle. A hotplate and water bath was used to maintain the
reaction solution temperature at ca. 30 oC. AlI of the monomer was added by syringe to
the stirring reaction solution through the septum on the addition buret. This septum was
then replaced with the catalyst storage flask. The addition buret was evacuated and
purged and four additions of 25 mL of the catalyst solution were made to the reaction
fiask, through the addition buret, quickly (ca. 2 min for each addition) at lime O. 0.5, 1.0.
and 1.5 h. The stirring rate of the reaction solution during the addition of catalyst was ca.
450 rpm. After the catalyst additions, the polymerization solution was stirred for another
2·20
•
•
•
Synthesis & Characrerizarion
13-15 h, under a continuous nitrogen fiow, and with the temperature maintained between
25 and 30 oC, as indicated hy the RTD probe.
The polymerization reaction was terminated by the slow addition, with a syringe,
of 54 mL of anhydrous ethanol, diluted approximately three-fold with anhydrous diethyl
ether, to the reaction mixture through the septum on the fifth neck. The mixture
containing the insoluble polymer product was allowed to stir for another 30 min.
2.2.5 Catalyst Removal
The product slurry was iniùally filtered through a Polywipe ® cloth, and washed
with 200-500 mL of anhydrous diethyl ether. The polymer was filtered again using
Whatrnan No. 1 filter paper and washed twice with 1000 mL of ether. The wet polymer
was added to a beaker containing 700 mL of reagent grade methanoJ. This slurry was
placed into a Waring Blender, using small portions at a time, and the particle size was
reduced until a very fine, almost milky, suspension was attained. 1be suspension was
added to a beaker containing 80 mL of a 10 % (wIv) hydrochloric acid/methanol solution.
The mixture was left to stir very slowly in the fume hood for 1.5 h. The acidified polymer
mixture was then decanted and ',vashed with fresh reagent grade methanol repeatedly, until
wet pH paper indicated a neutral pH. The product was finally filtered with Whatrnan
No. 1 filter paper, placed in a vacuum oven at room temperature, and dried to a constant
weight of 76 g.
2.2.6 separation of Ataetie and Isotaetie Poly(epiehlorohydrin)
The atactic PECH is soluble in hot acetone and it was separated from the isotactic
fraction in the following manner. The unfraction:\ted polymer was placed into a 5 L
Erlenmeyer flask which was fitted with a mechanical stirrer and a glass shaft with a Teflon
blade. Four Hters of reagent grade acetone was added to the flask containing the polymer,
and the slurry was stirred genüy overnight
2-21
Ct
Synthesis & Clwracterharicm
The polymer slurry was then filtered. using Whatman No. 1 tilter paper anù Ihe
filter cake was added 10 4 L of fresh acetone and slirred for anOlher 2 h. The polymer was
again filtered, and the acetone insoluble fraction was placed in a 5 L :I-neck round-hottom
flask containing 4 L of fresh acetone. The t1ask was equipped with a condenser with a
nitrogen flow over-top. The flask was immersed in a large waler hath which maimained
the solution at reflux temperature for 4 h. The polymer mixture was then allowed to cool
to room temperature overnight, with stirring. The polymer was then tiltered, set to rel1ux
again in fresh acetone, and the procedure was repeated. The isolated isotactic polymer
was washed with ca. 2 L of fresh acetone, filtered, dried, and stored under vacuum al
rrom temperature for at least 2 weeks before use. The tï,'.Ù yield of i-poly(R.S
epichlorohydrin) was 22 g (ca. 18 % of initial monomer).
2.2.7 Pclymerization of Optically Pure Monomers
Polymerization of each of the optically pure monomers was carried out under the
same conditions described above, with the reactions scaled down by a factor of live. The
reaction vessel used was a 500 mL 4-neck round-bottom flask. The catalyst was added in
three aliquots of 5 mL. Five grams of the optically pure epichlorohydrin monomer was
polymerized in each case. At the end of the reaction. the polym...r, an insoluble powder,
collected at the bouom of the reaction flask. The residual catalyst was removed according
to the procedure outlined in Section 2.2.3. The isolated polymer was used without
fractionation. The filtrate was evaporated to dryness and no residual solid or ftim was lefl
behind. (The yield was ca. 90 % of initial monomer).
2.2.8 Dechlorination of Poly(epichlorohydrin)
Bach of poly(R-epichlorohydrin) (PRECH), poly(S-epichiorohydrin) (PSECH). and
the isotactic fraction of poly(R.S-epichiorohydrin) (i-PRSECH) was dechlorinated by
reaction with LiAlH4• To ensure identical experimental conditions for each dcchlorinalion
2-22
Synthesis & Characterization
reaction, the three reactions were performed simultaneously. Using a top loading
analytical balance in~ide a glove bag in the fume hood, 3 aliquol~ of 1.03 g (a 2.5 molar
cxcess of aluminum to chlorine) of dry LiAlH4 were weighed and placed into clean, dry
100 mL 3-neck round-boltom flasks, equipped with magnetic stirring bars, and stoppered
with a rubbcr septum. To each flask was added ca. 45 mL of freshly distilled THF under
nitrogen using a double-tipped needle. Each flask was equipped with a condenser with a
nitrogen flow over top. AlI of the reaction flasks were placed into the same water bath
maintained at ca. 50 oC. The LiAlH4 was allowed to disperse by stirring each mixture.
before adding the polymer (1.0 g) to the reactio.. flask quickly. with a positive nitrogen
pressure. The reaction was allowed to proceed for three days.
To deactivate the excess LiAlH4• the hot water bath was first replaced with an ice
water bath. Then a solution of 16 mL of anhydrous ethanol in 50 mL of freshly distilled
THF was added dropwise to the reaction mixture. using a syringe, through a septum on
the reaction flask. The reaction mixture remained stirring for 30 min to 1 h and then was
slowly heated to room temperature.
Warrn. freshly distilled THF (ca. 25 mL) was added to the gray heterogeneoU5
reaction mixture which was then fùtered using Whatman No. 1 filter paper. The clay-like
filter cake was allowed to dry in air. The fùtrate was filtered again using high pressure and
a Teflon membrane fùter to remove residual insoluble matter. The filtrate was initially
placed on the rotary evaporator to reduce the volume of solution in the flask to a
minimum. The flask containing the product was then placed on a vacuum line, and heated
gently with a heat gun under vacuum. A clear, polymer film could easily be peeled l'rom
the sides of the flask with tweezers.
The film of poly(propylene oxide) was placed in a 25 mL l-neck round-bottom
flask. and HPLC grade benzene was added. The solution was freeze-dried to yield a white
powder. This polymer was dried and stored under vacuum at room temperature for a
minimum of two weeks before use.
2-23
•Synrhesis & Characterization
2.2.9 Preparation of Polyenantiomer Blends
2.2.9.1 Poly(epichlorohydrin)
Blends of different ratios of PSECH and PRECH were made hy di:.solving Ihe
appropriale amounts of each polymer in THF al 50 oC wilh stirring for~everJl hours
under a nitrogen atrnosphere. Fresh, cold anhydrous diethyl elher was used 10 prccipilale
lhe blend l'rom the 1 % (w/v) polymer solutions with vigorous stirring. The polymers
were dried \mder vacuum al room lemperalure for a minimum of IWO week.~ prior 10
analy.;is.
2.2.9.2 Poly(propylene oxide)
An eq'Jirnolar blend of poly(R-propylene oxide) (PRPrO) and poly(S-propylene
oxide) (PSPrO) was made by dissolving equal amounts of the polymers in HPLC grade
benzene. The solution was stirred al room temperalure for 1 h before immersing the l1ask
in a liquid nilrogen bath and removing the solvenl by freeze-drying 10 a dry, white mass of
polymer which was slored under vacuum for a minimum of 2 weeks prior 10 use.
2.3 Polymer Characterization
2.3.1 Gel Permeation Chromatography (GPC)
The weight average molecular weights of PREeH, PSECH, and i-PRSECH. werc
determined by gel permeation chromatography "1 50 oC, using a Varian D6·600 GPC in
tandem with a Varian (Model RI-4) refractive index deteclor. An injection volume of 50
ILL of warm 0.05 % (w/v) polymer solution in THF was used, with a flow raIe of 1.0
mUmin. The weight average molecular weights were found to be approxirnately 450 000
for ail of the polymers, as interpreted againsl polystyrene st:mdards and application of a
2-24
•
•
•
Synthesis & Characterhation
Q·faClOr of 25. The polymers were found to have a relaùvely narrow molecular weight
distribuùon with a polydispersity of 1.6 for both PRECH and PSECH and lA for the
isolaled isotactic fracùon of PRSECH.
The weight average molecular weights of the corresponding poly(propylene oxide)
polymers were also determined by GPC analysis at room temperature in chloroform. Both
PSPrO and PRPrO were found to have a broader distribution of molecular weights with
weight average molecular weights of approximately 110 000. If the dechlorinaùon
reaclion of PECH occurred withoUI any cleavage of the polymer chains, a 1.5-fold
reducùon in molecular weight would he expected in the resulting PPrO, arising simply
from the substitution of chlorine with hydrogen.')tellar28 has reponed a 3.5-fold
reducùon in molecular weight upon dechlorination of PECH 10 PPrO. The additional
decrease in molecular weight was accounted for by a cleavage of the polymer chains, due
10 the reaction conditions: dissolution (If the polymer in tetrahydrofuran at 50 oC for a
period of 50 h. In the present case, the dechlorination of PECH occurs with
approximately a 4-fold reduction in molecular weight of the resulting polymer. The
cleavage of the chains is deemed responsible for the observed broadening of the molecular
weighl distribution of the original polymer.
2.3.2 lSC Nuclear Magnetic Resonance Spectroscopy (lSC NMR)
2.3.2.1 Poly(epichlorohydrin)
The high molecular weight poly(epichlorohydrin) (PECH) is insoluble in common
solvents at room temperature, therefore NMR analysis of the polymer was performed at
elevated temperatures. Figure 2.3(a) contains the 13C NMR spectrum of the
unfractionated poly(R,S-epichiorohydrin) crude, filtered product after one washing with
cold acetone. The spectrum of a 5-10 % (w/v) polymer solution in deuterated chloroform
was recorded on a Varian XL-300 instrument at 50 oC. Figure 2.3(b) contains
2-25
•Synrhesis & Ch(/racleri~ati{)n
(a) 21
(b)
J 1 i • 1 il' 1 il' 1 i i • 1 il, 1 • 1 i i i 1
81 80 79 78 77 76 75 74 73 72 71 70 69 68
Chemical Shift Il (ppm)
2 1
Figure 2.3
~\ \ ~• i , i i i i i , 1 i 'i i i i i , i 1 i i' i 1 i i" 1 i i i' i" i' i i 1 i i 1 i i i il' i i, 1
79 78 77 76 75 74 73 72 71 70 69
Chemica! Shift li (ppm)
\3C NMR spectra of the backbone carbons of (a) the crude unfraclionated
PRSECH after one cold acelone wash, in COCll al 50 oC and (b) lhe
atactic PRSECH cold acetone soluble fraction in CDCll al 50 oC. showing
backbone carbon region only.
2·26
•
•
Synrhesis & Characrerizarion
the IlC NMR spectrum of the atactic fraction of the poly(R.S-epichlorohydrin) polymer
product, soluble in cold acetone. This spectrum was collected with a Varian UnitY 500
instrument at 50 oC using a 10 % (w/v) solution of polymer in deuterated chloroform at
50 oc. Figures 2.4 and 2.5 contain the I3C NMR spectra and expanded methylene carbon
regions of the fmal isotactic fraction of poly(R.S-epichiorohydrin) (i-PRSECH), and the
unfractionated optically pure PRECH polymer, respcctively. These spectra were recorded
on a Varian Unity 500 NMR spectrometer operating at 125 MHz using 5-10 % (w/v)
polymer solutions prepared with deuterated dimethyl sulfoxide at 70 oC. TIle pulse angle
for these experiments was 70° with ais acquisition rime and a 2 s delay time, and the
number of transients was equal to 16 000.
Except for a 10wer concentration and 10wer temperature, the operating conditions
employed in obtaining the spectra of i-PRSECH and PRECH, were the same as those used
by Cheng and Smith29 in their investigation of regioregular poly(R.S-epichiorohydrin),
synthesized using the same method as in the present study, Le., the Vandenberg method.26
Scheme 4 is a schemadc diagram containing their fine structure resonance assignments for
the stereorandom, regioregular fraction.
Scheme4
ml
CH2(69 ppm)
1
CH(78 ppm)
nvnr-f-,rrmr+rm
invn irm
imr
ri
CH3(43 ppm)
1rr (sin~t)
•This interpretational scheme was used to make the peak assignments in the PRECH and
i-PRSECH spcctra.
The absence of inverted sequences (regioirregularities resulting from head-to-head
or tail-to-tail additions of monomer) in the PRECH and the i-PRSECH polymers is
consistent with previous reports28-30 and confirms that during the polymerization of
2-27
•SynrhesÎs & CharacterhlllÎOn
CH,
rn-,mm nn mr rr
68.9 68.8 68.7 68.6 68.5 •
i 1 i i 1 i i 1 i i 1 i i 1 i i 1 i i i i i 1 i i 1 i
80 75 70 65 60 55 50 45 40
Chemical Shift B(ppm)
Figure 2.4 13C NMR spectrurn and expanded methylene region of i-PRSECH fraction
in deuteraled DMSO at 50 oC.
2-28
•Synthesis & Characterization
'I1i"1 Iii '1 il i '1 El "Ii;68.9 68.8 68.7 68.6 6:l.5
"i'I"liliil'·i'i'·i'I"i'I""J'i"I'ii'l'
40455055606570
Chemical Shift Ô (ppm)
7580
Figure 2.5 13C NMR spectrum and expanded methylene region of PRECH in
deuterated DMSO al 50 oC.
2-29
•Synrhesis & CharaCleri:mion
optically active epichlorohydrin with Et)A1·H20 according to the Vandenhcrg methoù.26
only one of the C-O bonds of the asymmetric monomer ring is subject to deavage. Based
on the relative peak areas at 68.55 ppm. 68.63 ppm, and 68.73 ppm to that at 68.711 ppm,
the stereoirregulariùes manifested in the spectrurn of i-PRSECH make up no more than
3 % of the polymer chain. These are attributed to junctions or defect sites between long
stereoregular sequences, and not to a separate atactic chain fraction. Esscntially, the
i-PRSECH polymer can he described as a stereoblock polymer. Assuming a r,mdom
occurrence of defect sites a10ng the polymer chain, the average length of a stereosequence
is ca. 33 monomer units.
2.3.2.2 Poly(propylene oxide)
Figure 2.6 contains the IlC NMR spectrurn of the isotacÙc poly(R,S-propylene
oxide} (i-PRSPrO) obtained by the dechlorinaùon of the i-PRSECH polymer. The
spectrurn was obtained with a Varian XL 300 NMR spectrometer operaùng at 50 oC,
using a 20-30 % (w/v) polymer solution in deuterated chloroform. The carbon resonances
of the i-PRSPrO polymer appear in the sarne order as those of the PECH polymers, with
the pendent methylene carbon resonance the furthest downfield, at about 17 ppm, the
backbone methylene carbon resonance appearing at about 73 ppm, and that of the methine
carbon at about 75 ppm.
Figure 2.7 contains the IlC NMR spectrum of the polymer product of incomplete
dechlorination of poly(S-propylene oxide} (PSPrO). The reacùon of PSECH with LiAlH4
was stopped after 5 h, and the resulting polymer was a clear, sticky paste which, under the
polarized light microscope, did not show any signs of birefringence upon slow cooling
l'rom 150 oC. According to the relation of reacùon lime and degree of dechlorination
presented in Figure 2.1 for the PECH elastomer, roughly 60 % of the original chlorine
remains in the polymer. The spectrurn was recorded with a Varian Unity SOO
Specl;'ometer using ca. 20 % (w/v) soluùons in deuterated chloroform at 50 oC. ln Figure
2-30
1
2
Synthesis & Characterization
3
1 i i 1 i i 1 i i 1 i i 1 i i 1 i 1 i
80 70 60 50 40 30 20
Chernical Shift Ô (ppm)
Figure 2.6 13C NMR spectrwn of i-PRSPrO in deuleraled chlorofonn al roomtemperalure.
2-31
•SyntiJesis & CiJaracreri:arion
(c)
i 1 i i i 1 i i
17.5 17.0
(a)
Iii i i i i 1 i
(b)
1 1
45
i 1 i
, 1 i
44
i 1 i
i 1
43
i 1
80 78 76 74 72
Chemical Shift B(ppm)
70 68
Figure 2.7 \3C NMR spectrum of the product of the incomplete dechlorination
reaction of LiAlH4 on PSECH, in CDCl3 at room temperature: (a) PSECH
and PRPrO backbone carbon region, (b) PSECH pendent carbon region,
and (c) PRPrO pendent carbon region.
2·32
Synthesis & Characterizotion
2.7(a) the main chain carbon region is expanded. which shows the presence of both PPrO
and PECH carbons. a~ weU as resonances corresponding to PPrO and PECH neighboring
cacbons. The insets (b) and (c) contain the pendent methyl carbon spectral regions of
PECH and PPrO. respectively. It is interesting to note that a1though at this point the
polymer is a random copolymer and evidently amorphous. ail of the methyl and
chloromethyl pendent groups are in an isotactic configuration. Dimonie et a1.31 have
reported the copolymerization of racemic propylene oxide and epichlorohydrin monomers
using the Et3Al·H20 catalyst The copolymers were not reported as being crystalline.
2.3.3 Polarimetry
The specific angles of rotation of the POO polymers were deterrnined using a
JASCO DIP-140 Digital Polarimeter, operating at room temperature. and using the
sodium D line (Â. = 589 nm). Polymer solutions were prepared as 1.00 % (w/v) in
spectrophotometric grade chloroforrn. The optical ceU path length was 50 mm.
Table 2.1 Optical Rotation Angles of Poly(propylene oxide) Sarnples
__--"-po;;.:1"'ym=er:...- -'~2o CHCI3
PSPrO +24±0.3°
PRPrO -24 ± 0.3 0
i-PRSPrO +9 ± 0.3 0
The fmdings in Table 2.1 agree with th,; reported values of +25 0 and +24 0 for
poly(1)-(+)-(propylene oxide) under the same conditions.16 Since reduction of the PECH
occurs with retention of configuration of the chiral chains, the optical purity observed in
the resultant PPrO polymers is a direct reflection of the optical purity in the precursor
PECH polymers. Thus, the PRECH, PSECH, PRPrO and PSPrO used in this thesis are
essentially ail optically pure polyenantiomers. A degree of optical activity (ca. 37 %
optical purity) is not unexpected for the stereoblock polymer. Optical rotation angles of
2-33
•SynriJesis & CI/(jracteri~arioll
± 5 0 have been found for the crystalline, isotacùc fraction of PRSPrO synthcsized l'rom
racemic monomer.16
2.3.4 Fourier Transform Infrared Spectroscopy (FT-IR)
A Perkin Elmer Fourier Transform infrared spectrometer was employed in the
general characterization of the polyethers. The PECH samplcs were prepared hy simply
melt-pressing a polymer crumb between Teflon sheeL~ on a hotplate at the appropriate
temperature. A secùon of uniform thickness (ca. 20 Ilm) was used for acquiring the
spectrum. The PPrO samples were prepared by placing a few drops of a conccntrated
chloroform solution directly onto the salt plate and evaporating to dryness. Figure 2.8(a)
contains the infrared spectrum of the equimolar blend of PSECH and PRECH
polyenantiomers, and in (b) the spectrum of the stereoblock i-PRSPrO is presented.
The representative spectra can he compared for structural dillerences. in particular
for the presence of chlorine. Ishida et al.32 reported the IR spcctra, without peak
assignments, for the atactic and isotactic fractions of poly(R,S-epichlorohydrin) and
poly(R.S-propylene oxide), each synthesized from the respective racemic monomer using
the Pruitt and Baggett catalyst.15 AIl of the peaks in the reported PECH and PPrO spectra
can he found in the present spectra in Figures 2.8(a) and (b), respecùvely, with identiL'al
peak positions and relative intensities.
Kawasaki et al.33 have reported the detailed assignments of the IR absorptions of
both the atactic and isotactir, fractions of poly(R.S-propylene oxide), prepared using a
stereoselective catalyst. These assignments were used to idenùfy the peaks in the PPrO
spectrum. The peaks which appear only in the spectrum of PECH are indicated by an
arrow, in Figure 2.8(a). Both spectra show a CHz peak at about 1480 cm'); CH bending
peaks in the region of 1350-1330 cm-l but with differing intensiùes; CHz wagging and
twisting frequencies between 1200 and 1300 cm-1; and the characteristic broad ether peak
centered around 1100 cm-l • By contrast, the CH3 peaks at about 1450 cm· l , 1380 cm-)
2-34
•Synthesis & Characterization
a
t
t
t
UCH
1 1 1Ob CH, CHz
lWin
tb
1CH,
UCH,rocIdua
C·().C
1500 1400 1300 1200 1100 1000 900 800 700
Wavenumber (cm-1)
Figure 2.8 Fr-IR spectra of (a) the equimolar blend of PSECH and PRECH
polyenantiomers and (b) i-PRSPrO.
2-35
•Synrhesis & Characleri:arion
and 900 cm-I appear only after dechlorination in the PPrO spectrum. Most notahly, the
peaks in the characteristic C-Cl slrelching region (800 to 650 cm-I) in the PECH spectrum
are absent from the spectrum of PPrO. The same is true for the spectra 'JI" the individual
polyenanùomers, PRECH and PSECH (not shown here).
2.3.5 Wide-Angle X-Ray (WAX) Diffraction
X-ray powder diffracùon patterns and liber diffracùon patterns were ohtained with
a fiat fùm Warhus camera using Ni-fùtered Cu-Ka radiaùon, under vacuum. The X-ray
beam was generated by a Philips PW 1730 X-ray generator (40 kV and 20 mA). The
camera length and exposure tÎme was as stated in the Results Section. Sodium l1uoride''''~''1
W;lS used as the calibration standard, and a drop of a saturated solution was placed onto
the mounted sample and evaporated to dryness prior to X-ray analysis.
The samples were prepared for X-ray analysis by tirst pressing the polymer
belWeen two microscope glass slides lined with a Teflon sheet on a hotplate at the
appropriate temperature (175 oC for PECH and ISO oC for PPrO) unùl the sample mclted.
The polymer assembly was then quickly transferred to a Mettler FP52 hotstage at the
same temperatule with a nitrogen stream passing over the sampIe. The polymer assembly
was held isothennally in the melt for 20 min and then cooled slowly to the crystallizaùon
temperature (60 oC for PECH and 30 oC for PPrO) where it remained for 1 h. The IiIm
was then removed from belWeen the microscoDe slides and was placed in a vacuum oven
at the annealing temperature (80 oC for PECH and 50 oC for PPrO) for 15 h. The non
oriented films were of the order of 50 - 100 IJ.m thick.
Oriented samples of the PECH polymers were prepared for analysis by heatÎng a
crumb of the polymer on a hotplate at ca. 100 oC, while grasping the sample with tweezers
and stretching until a narrow, librous strip of ca. 50 mm in length was attained. The liber
was mounted onto the sample holder with the stretching direction roughly perpendicular
to the direcùon of the beam.
2-36
•
•
•
Synthesis & Characrerizarion
In Figure 2.9 the X-ray powder diffraction patterns of the stereoblock i-PRSECH
(a), the PRECH polyenantiomer (b), and the equimolar polyenantiomer blend (c), are
displayed. The pattern of the PSECH sample (not shown here) is indistinguishahle l'rom
that of PRECH. To a frrst approximation, the series of diffraction rings for each of the
polymers are identical. In (a) a narrow diffraction line width superimposed on the
background intensity indicates a degree of order in the i-PRSECH crysta1s which is
comparable to that of the polyenantiomers, and their blend.
The liber diffraction patterns of the PRECH and i-PRSECH samples are contained
in Figures 2. IO(a) and (b) respectively. Again, since the diffraction patterns of the
PRECH, PSECH, and their equimolar blend are indistinguishable, only one pattern is
shown. The stereoblock i-PRSECH polymer exhibits the same pattern except for the
apparent düferences in the relative intensities of the reflections. The peculiar intensity
pattern was reproducible. The representative PRECH liber diffraction pattern was
digitized and the d-spacings were measured usir.g JAVA (Jandel Scientific) video image
analysis software. The values of the measured d-spacings with the relative intensities were
found to correspond to within two significant ligures with the 15 most intense reflections
indexed and reported by Richards34 according to an orthorhombic unit cell of dimensions a
=12.14 Â, b =4.90 Â, and c =7.07 Â (liber axis) for isotactic poly(R,S-epichlorohydrin).
From the PRECH liber pattern in Figure 2.10(a), the liber repeat distance is determined to
be 7.02 ±0.01 Â. This value is closer to the value of 7.05 ± 1 Â reported by Ishida32 than
the above value reported by Richards.
The powder diffraction pattern of the stereoblock i-PRSPrO is presented in Figure
2.1 I. The d-spacings of the most intense rings, measured using a microcaliper. were
compared with the three most intense reflections of i-PRSPrO reported by Ishida.32 In
decreasing intensity, these are 4.24 Â. 5.18 Â, and 2.78 Â. in excellent agreement with the
reported values of 4.23 Â. 5.17 Â. and 2.78 Â. The powder diffraction patterns of
PSPrO. PRPrO and i-PRSPrO were indistinguishable. Based on the above fmdings, the
2-37
•
•
Synth~sis '" Charactuization
a b c
•Figure 2.9 Wide-angle X-ray powder diffraction patterns of (al i-PRSECH.
(bl PREeH. and (cl the equimolar polyenantiomer blend. Diffraction spots
are due to the calibration standard (exposure lime =2-4 hl.
2-38
SynrhesÎs & CharacrerÎwrÎon
b
Figure 2.10 Widc-angle X-ray liber diffraction patterns of (a) PRECH and
(b) i-PRSECH (exposure ùme ::18 hl.
2-39
•Synrhe.ûs & CllIIraererbll;Oll
Figure 2.11 Wide-angle X-ray powder diffraction pattern of i-PRSPrO (exposure lime
=4h).
2-40
•
•
•
Synthesis & Characterization
polyenantiomers, their equimolar blend, and the stereoblock form of PECH crystallize
from the melt with the same crystal structure. For PPrO, it is also found that the
polyenantiomers and the stereoblock form the same crystal structure from the melt
2.3.6 Differentiai Scanning Calorimetry (DSC)
The basic thermal behavior of each polymer was studied by differential scanning
calorimetry using a Perkin-Elmer DSC-7 instrument, calibrated for temperature and peak
arca using indium, octadecane, anl1 octane high quality standards (Aldrich) and operating
under a nitrogen purge gas at ail times.
2.3.6.1 Glass Transition Temperature (Tg)
The glass transition temperature (Tg) was determined for each of the PECH
polymers: PRECH, PSECH, the equimolar polyenantiomer blend, and i-PRSECH, and for
thc PPrO samph:, i-PRSPrO. To determine the Tg, a 5 mg sample of polymer was lirst
melted at the appropriate temperature (175 oC for PECH and 150 oC for PPrO) with the
OSC for 15 min. The sample was then quench-cooled at a nominal rate of ISO oC/min to
a temperature weil below the expected glass transition temperature (-40 oC for PECH and
-100°C for PPrO), equilibrated, and finally heated at a scanning rate of 10 oC/min. The
point of inllection in the thermogram transition is taken as the temperature of the Tg.
Figure 2.12 contains the heating thermograms used to determine the value of the
Tg for each of the PECH samples. Analysis of both the optically pure polyenantiomers and
the stereoblock polymer results in a common Tg value of -26 oC. Janeczek et aI.3S
reported a value of -27 oC, determined by OSC, for isotactic poly(R.S-epichiorohydrin).
As the corresponding OSC thermogram of the i-PRSPrO sample in Figure 2.13 illustrates,
the effeet of substitution of the ehlorine atom on PECH with a hydrogen atom lowers the
Tg significantly, to a value of -65 oC for the resulting more fleXible i-PPrO polyether. It is
weil understood that the glass transition temperature depends upon chain flexibility and
2-41
•Svnthesis & Characreri:arioll
o
~
a ............-._._._._._._._._.--'-""--'-'-'-'
-b
,,,,,,--.-..._--- -----,,/
C .'.....---,'-'",,-
d ./ ----- ------'.._----• , 1 • 1 • 1 l ' • ! • • , • 1 • • • , 1 ••
-40 -3." -30 -25 -20 -15 -10 -5 oTemperature (OC)
Figure 2.12 DSC thermograms of melt-quenched (a) PRECH, (b) PSECH, (c) the
equimolar polyenantiomer blend, and (d) i-PRSECH showing the Tg (all
heating rates =10 oC/min).
2-42
•Synthesis & Characterization
o
~
I-..."""T'""T"""""""'I-"""'"• """T'""T"""""""'I"""T'"-..-,.........1-,.........-r""T""",.........-.-:.'.'. -.- 1 • •
-100 -90 -80 -70 -60
Temperature (OC)
-50 -40
Figure 2.13 Dse thennogram of melt-quenched i-PRSPrO showing the Tg (heating rate= 20 oC/min).
2·43
•Symhesis & CharaClerizalion
polymer polarity. There are a number of values for the Tg of i-POO repoI1ed in the
literature. These range from -75 °C36 to -60 °C37.
2.3.6.2 Equilibrium Melting Temperature (Tm0)
To determine the equilibrium melting temperature, the sample preparation involved
fust melting samples of less than 2 mg for 15-20 min in the DSC instrument. then cooling
at a nominal rate of 150 oC/min to the appropriate crystal1ization temperature. Typically.
the samples were held isothermal for ca. 15 and 24 h for the POO and the PECH
polymers, respectively. The thennograms were recorded using a scanning rate of
20 oC/min. A fresh polymer sample was prepared for each crystallization temperature.
The equilibrium melting temperature (Tm0) of a crystalline polymer is defined as
the melting temperature of a perfect crystal fonned by infinite molecular weight chains.3g.39
In general, the observed melting point depression in lamellar crystallites is due to the large
surface to volume ratio. This ratio decreases with increasing crystallization temperalUre
(TJ as relatively more perfect, thicker lamellae are fonned at higher (TJ. From Hoffman
and Weeks,40 the relation between the observed melting point (Tm) and Tc is given by:
T = T 0(1_.!.\J-r Tcm m y y (2.1)
•
where y is the lamellar thickening factor.41 The Tm° of each of the PECH and PPrO
polymers was determined by extrapolating the data in the Hoffman-Weeks40 (H-W) plots
of Tm as a function of Tc until intersection with the Tm =Tc line.
Figure 2.14(a) and (b) contain lypical DSC traces of PRECH and PSPrO,
respectively, used in the construction of the (H-W) plots. Figure 2.15 contains the H-W
plot for i-PRSECH, PRECH, PSECH, and the equimolar polyenantiomer blend. The co
linearity of the melting point data leads to a common extrapolated value of Tm° = 138
2-44
•Synthesis & Characterizarion
t
(a) (b)
rJ
\
1 1 1 1 1 1
110 115 120 125 55 60 65 70 75
Temperature (OC)
Figure 2.14 Typical OSC heating thennograms used in the construction of the
Hoffman-Weeks plots of (a) PSECH isothennally crystallized for 24 h and
(b) PSPrO isothennally crystallized for 15 h, both at an undercooling of
30 oC, .'nd heated at 20 oC/min.
2-45
•150
140,-...Ua'-" 130~::l~ 1201-0Q)0..E~ 1100.0c:.- 100...
Q)
~90
80 .
Synrhesis & Charactl'ri:arioll
.'"
.'
.'
aBD ....
~ 8~.....,""~Tm = Tc
"
80 90 100 110 120 BO 140 150
Crystallization Temperature (OC)
Figure 2.15 Hoffman-Weeks plot of the observed melting temperature as a funclion of
crystallization temperature for optically active PRECH 'and PSECH (e).
the equimolar polyenantiomer blend (.). and the stereoblock i-PRSECH
(.â.). The hollow syrnbols represent the corresponding lower melting
endotherrn maxima.
2-46
•
•
•
Synthesis & Characterizarion
± 1°C, (411 ± 1 K) for ail of the PECH samples. Kambara and Takahashi42 reponed a
similar value of 135 oC for poly(R,S-epichlorohydrin). In Figure 2.16, the H-W plot for
the PSPrO polyt:nantiomer is presented. Extrapolation of the data yields a Tm° value of
82 ± 1°C, (355 ± 1 K). Samples of PRPrO, i-PRSPrO, and the equimolar blend of the
polyenantiomers were crystallized at selected temperarures in the range of 5D-60 oC for
15 h and had an observed melting temperature which fell on the PSPrO H-W line. On the
basis of these results, the value of 82 ± 1 oC (355 ± 1 K) was taken as the Tm° for ail of
the PPrO polymers. Booth et al.43 reponed aTm0 of 82 oC for the isotactic crystalline
fraction ofpoly(R,S-propylene oxide) with a mole fraction ofisotactic diads of98 %.
The results of common extrapolated Tm0 values for thc polyenantiomers, the
polyenantiomer equimolar blend, and the stereoblock forms of each polyether implies that
the different polymer forms have same basic crystal structure and, at the high
crystallization temperatures used in this study, the most perfect crystallites in each sarnple
possess similar degrees of order.
2.4 Summary
The optically active R and S forms of epichlorohydrin were polymerized. using
Et3A1.~0 (1:0.6) catalyst, to the corresponding regio- and stereoregular high molecular
weight polymers (PRECH and PSECH). The polymers were used without fractionation.
A regio- and stereoregular high molecular weight fraction (i-PRSECH) was obtained from
the polymerization of the racemic monomer under simi1ar conditions. The different forms
of the poly(epichlorohyclrin) (PECH) polymers were quantitatively convenP.d to the
corresponding forms of poly(propylene oxide) (PPrO), using LiAlH4 reducing agent, with
the chiral centers remaining intact. Subsequent ana1ysis of the PPrO polymers by Fr-IR
and 13C NMR spectroscopy did not deteet PECH. For each of PECH and PPrO,
2-47
•Synthesis & Characterization
90 lû 80°'-'e~ 70
~E-- 600()c::.-.....
Q)~ 50
.'.'
......,....,T =Trn c
50 60 70 80 90
Crystallization Temperature (OC)
40 +-.......,.--r-__r____,~_r___r-_r___r____,__l
40
Figure 2.16 Hoffman-Weeks plot of observed rnelting ternperature as a function of
crystallization ternperature for i-PRSPrO.
2-48
•
•
•
Synrlzesis & Clzaracreriuuion
the different fonns of the polyrners possess the same molecular weight and molecular
weight disoibution, which was unimodal for PECH and slightly bimodal for PPIO.
On the basis of the 13C NMR speclrOScopy results, it is concluded that i·PRSECH,
and consequently its derivative i·PRSPrO, are each stereoblock polyrners with
characteristic long stereoregular sequences. Optical rotation measurements of PSPrO and
PRPrO clearly show that the polyrners are indeed optically prue polyenantiomers, and
therefore it is concluded that PRECH and PSECH are also optically pure polyenantiomers.
On the basis of the WAX diffraction patterns, i-PRSECH possesses a degree of
order in the crystals which is comparable to that in the polyenantiomers. For each of
PECH and PPrO, the equilibrium melting temperature (Tm0), as determined by Hoffrnan
Weeks analysis, is the same for the optically pure polyenantiomers, their equimolar blend,
and the stereoblock fonns.
2.5 References
(1) Billingham, N. C. In Developments in Polymerization-l; Haward, R. N., Ed.;
Applied Science Publishers: London, 1979; Chapter 4.
(2) Colclough, R. C.; Gee, G. J. Polym. Sei. 1959,34, 153.
(3) Sakata, R.; Tsruta, T. Makromol. C/zem. 1960,40, 64.
(4) Vandenberg, E. J. J. Polym. Sei., A-I 1969, 7, 525.
(5) See for example,(a) Principles of Polymer Chemistry, Ravve, A.; Plenum: New
York, 1995; Chapter 4: Ring Opening Polymerizations. (b) Prineiples of
Polymerization, Odian, G. C.; Wl1ey: New York, 1991. (c) Ring Opening
Polymerization: Kineties, Meelzanisms and Synt/zesis, McGrath, J.E., Ed.; ACS
Symp. Ser. 1985.
(6) Sigwalt, P. Pure Appl. Chem. 1976,48,257.
2·49
•Synrhesis & Characterization
(7) Spa~sky. N. Polym Prepr. (Am. Chem. Soc.. Div. Polym. Chem.) 1977. 18.76.
(8) Wurtz. M. A. Ann. Chim Phys. 1863.3,427.
(9) Levene. P. A.; Walti. A. J. Biol. Chem 1927. 73.263; 75. 325.
(10) Staudinger. H.; Lohmann. H. Ann. Chim 1933.505.41.
(II) Natla. G. J. Polym. Sei. 1955. 16. 143.
(12) Natta. G.; Pino. P.; Corradini. P.; Danusso. F.; Mantiea. E.; Mazzanti. G.;
Mor~glio. G. J. Am Chem Soc. 1955. 77. 1700.
(13) Vandenberg E. 1. ln Polymer Seience and Technology; Priee. C. C., Vandenberg.
E. J.• Eds.; Plenum Press: New York, 1983; Vol. 19. pp 11-43.
(14) (a) Vandenberg, E. 1. V.S. Patent 3.058.963, 1962. (b) Vandenbt'rg, E. 1. V.S.
Paient 2,954,367, 1960.
(15) Pruitl, M. E.; Baggetl, J. M. V.S. Patent 2,706.181.1955.
(16) Priee, C. C.; Osgan, M. J. Am Chem So,. 1956, 78,4787.
(17) Osgan. M.; Priee, C. C. J. Polym. Sei. 1959,34,153.
(18) Vandenberg, E. J. J. Polym Sei.• Part C 1963. 207.
(19) (a) Vandenberg, E. 1. V.S. Patent 3.135,705, 1964. (h) Vandenberg. E. 1. V.S.
Patent 3,219,591,1965.
(20) Vandenberg, E. J. Polymer 1994, 35. 4933.
(21) Furukawa, J.; Tsruta. T.; Sakata, R.; Saegusa. T.; Kawasaki, A. Makromol. Chem
1959, 32, 90.
(22) Colclough, R. O.; Gee, G. Higginson, W. C. E.; Jackson, J. B.; Litt, M. J. Polym
Sei. 1959,39, 171.
(23) Vandenberg, E. J. J. Polym Sei. 1960,47. 486.
(24) Priee, C. C.; Fukutani, H. J. Polym Sei., A-ll968, 6, 2653.
(25) Vandenberg, E. J. Pure Appl. Chem 1976, 48, 295.
(26) Vandenberg, E. J. ln Macromolecular Synrheses; Bailey, W. J., Ed.; Wiley, New
York, 1972, Vol. 4, p 49.
2-50
Synthesis & C/raractt'ri:lllioll
(27) Vandenberg. E. J. ACS Symp. Ser. 1992.496.2.
(28) Steller. K. E. ACS Symp. Ser. 1975.6. 143.
(29) Cheng. H. N.: Smith. D. A. J. Appl. Polym. Sei. 1987.34.909.
(30) (a) Dworak. A. Makromol. C/rem. Rapid Commull. 1985.6.665. (bl Lindfors. K.
R.: Pan, S.: Dreyfuss, P. Macromolecules 1993. 26, 2919.
(31) Dimonie. V.: Donescu. D.; Gavai. I. Rev. Roum. C/rim. 1974. 19.125.
(32) Ishida, S. Bull. C/rem. Soc. Jpn. 1960.23.727.
(33) Kawasaki. A.: Furukawa. 1; Tsuruta. T.; Saegusa, T.: Kakogawa. G.: Sakala. R.
Polymer 1960, 1, 315.
(34) Richards, J. R. Ph.D. Thesis, University of Pennsylvania. Philadelphia. 1961. Diss.
Abstr. 1961,22, 1029.
(35) Janeczek, H.; Trzebicka. B.; Turska. E. Polym. Commun. 1987. 28, 123.
(36) Kuntz, I.; Cozewith. C.; Oakley. H. T.: Via. G.; White. H. T.: Wilchinsky, Z. W.
Macromolecules 1971,4,4.
(37) Faucher, 1. A. J. Polym. Sei. 1965. B-3, 143.
(38) Flory, P. J. J. Chem. Phys. 1949, 17, 223.
(39) Mandeikem. L. Crystallization ofPolymers; McGraw-Hill: New York. 1964.
(40) Hoffman. 1. D.; Weeks, J. J. J. Chem. Phys. 1962,37, 1723.
(41) Alamo, R G.: Viers, B. D.; Mandeikem, L. Macromolecules 1995, 28,3205.
(42) Kambara, S.; Takahashi. A. Makromol. Chem. 1963. 63, 89.
(43) Booth, C.; Dodgson, D. V.; Hillier, 1. H. Polymer 1970, Il. 11.
2-51
THERMAL BEHAVIOR
Coopter Three
3.1 Introduction:
Multiple Melting Behavior in Polymers
ln contrast to the sharp melting points of small molecules, macromolecules often
melt over a temperature range. The relatively broad melting region is indicative of a
spcctrum of crystallite" order within the solid phase. The presence of crystallite defects
depresses the melting temperature of the metastable crystallites by varying amounts.
Lamel1ar crystallite defects include chain ends and 100se folds on the surface, tie molecules
between lamel1ae, and defects of the internai structure. Since polymer lamel1ar crysta1lites
characteristically have relatively extensive surface areas, the contributions of the surface
defects to their metastability will consequently exceed any contributions from the internai
structure defects. 1 Such an array of crystallite perfection, crystallite sizes, and also a
broad molecular weight distribution can account for a broad, single-peak melting
endotherm observed by differential scanning calorimetry (OSe).
It is wel1 known that when a polymer is maintained at a temperature above its
crystallization temperature (Tc) but below its melting temperature (Tm)' structural
" 77le term crystal is used ta describe po/ymer single crysta/s and crysta//attice systems; tlzeterm crystallite is used ta describe po/ymer lamellae formed in bulk po/ymers.
3-1
•
•
Thermal Behm'ùlr
rearrangements can occur, bya proœss of annea/ing,2 In principk, the polyrner chains in
Ihe amorphous surface regions of the crystallite hecome mobile so that they can he dr.lwn
into the folded-chain crystallite structure, Annealing the polyrner sample can result in a
shift of the melting peak 10 a higher lemperalure and a narrowing of the distrihutions or
cryslallite order, as a result of changes in the lamellar thickness.2
Under certain thermal analysis conditions, many semicrystalline polyrners exhihit
more than one maximum in the melting endotherm. The so-called multiple melting
behavior can be representative of the fusion processes of material of differenl levc1s of
crystaI1ite perfection (order). These include multiple crystal lauice types, and multiple
morphologies, e.g. folded- and partially extended-chain crystallites, which exist in the
sample prior to the thermal analysis. An alternative interpretation is ouùined as follows.
In principle, when a polymer sample passes through the annealing temperature rcgion
during the thermal analysis scan, the conditions may permit the occurrence of structural
rearrangements typical of anneaIing behavior. In this case, the concurrent endothermic
and exothermic processes of the structural rearrangements may result in an ovcrall,
complex multiple peak appearance of the melting endotherm. The superposition of an
exothermic trough on a single peak endotherm can give the appearance of two melting
peaks. These interpretations are, in their simplest form, the two explanations which have
been used in the Iiterature to rationalize the observed multiple melting behavior common
to many polymers. An account of sorne of the early investigations of the phenomenon of
multiple melting in semicrystalline polyrners is presented below, with the focus on the
developmenl of the reorganization model. For clarity, the investigations cited in Sections
3.1.1.1-2 are not presented in chronological order bul grouped according 10
interpretation.
3-2
•Thermal Behavior
3.1.1 Historical Note
3.1.1.1 Polymer Single Crystals
In 1967 Bair el al.) investigated the multiple melting peak behavior exhihited in Ihe
DSe thermogram upon heating of polyethylene singl.: lamellae. The sample was cross
linkcd hy irradiation, thereby rendering the lamellar surface molecules immohile and
preventing lamellar perfection by thickening. After irradiation, the polymer exhibited a
single peak endotherm in the ose thermogram. The resuilS of this early study
demonstrated the potential of lamellar crystals to undergo sorne type of annealing process
during the ose scan, and ilS effect on the appearance of the melting thermogram.
Specifically, the process of reorganization in the initial ose scan was thought to involve
the melting, or partial melting, of the relatively disordered crystals, indicated by the lower
temperature peak, followed by recrystallization into a more perfect crystal which melted at
the temperature of the higher temperature peak.
3.1.1.2 Bulk Polymers
The notion of reorganization during a thermal analysis s'.:an was not quickly
adopted to account for the observed multiple melting behavior in bulk polymers. In the
initial reports, the situation was complicated by the apparent anomaly that the material
with the higher melting temperature could be converted into the lower melting
temperature material with the appropriate selection of scanning rates during the thermal
analysis.4-7 With increasing heating rate or increasing annealing temperature, the position
of the lower temperature peak increased and the endotherm became increasingly dominant
in magnitude, apparently at the expense of the higher temperature endotherm.
Yubayashi et a1.8 investigated the multiple rnelting peak behavior of annealed and
cold-drawn poly(ethylene terephthalate) (PET) by differential thermal analysis (OTA).
The high ternperature endotherm was thought to represent the rnelting of the initially
3-3
•
•
•
Thermal Behavior
present folded-chain crystallites which on pre-scaJlning thermal annealing and/or cold
drawing was partially converted to bundle-like crystallites which meIted at the lower
temperature. The bundle-like rnaterial was described as having only domains of extended
chain crystals so that overall it was less stable towsrds melting than the initially present
folded-chain material. Both the higher and the lower temperature endotherms were
thought to represent the melting of material present before the thermal analysis scan. This
slUdy did not consider the possibility of any structural changes taking place during the
instrument scanning process.
Bell and Dumbleton6 examined the multiple melting behavior of annealed and cold
drawn nylon-6,6 and isotactic polystyrene (i-PS) by performing DSe using varying heating
rate experiments. Once again, both melting endotherms were attributed to rnaterial
initially present in the sample. They added however, that the higher melting folded-chain
rnaterial was the thermadynamically{avored morphology and the lower melting bundle
like rnaterial was the /dnetically{avored morphology, but which as-produced was less
perfec!. The increase in temperature of the lower temperature melting peak with
increasing heating rate, was accounted for by the susceptibility of the bundle-like rnaterial
to thermal conductivity problems like superheating. Superheating of a sample is a
phenomenon in which the applied heating rate exceeds that rate at which the melt·solid
boundary can advance through the sample.2 The unmelted inner portion is consequendy
observed to melt at higher temperatures. This interpretation was maintained by Bell and
Murayama9 who performed similar DSe slUdies on PET.
The complex nature of the observed melting behavior was illustrated by the
conflicting views presented by different authors. For example, while Roberts' maintained
that any reorganization of PET crystallites must occur during the pre-thermal analysis
trealment as opposed to during the thermal analysis scan, he submitted a conffieting
assignment of the lower and higher temperature endotherms of PET. Roberts assened
that the melting of the folded-chain material initia1ly present was illustrated by the lower
3-4
•Thermal Behavior
temperature endotherm, and that this material could be converted by the pre-scanning
trcatment to partially chain-extended crystals, represented by the higher mdting
cndotherm. Similarly, Zachman and StuartlO examined PET by ose and wide-angle X-ray
scattering (VIAXS) techniques. Isothermally crystallized samples were annealed for
various lengths of time and subsequcntly quench-cooled and then examined. It was
discovcrcd that the density of an isothermally crystallized sample initially decreased at the
onset of annealing but then increased to exceed the original value. The higher temperature
endotherm represented the melting of crystalIites of greater perfection, produced from the
partial melting and recrystallization of the originalless perfeet crystallites. The possibility
of crystallite reorganization during the ose heating scan however, was not considered.
On the other hand, based on the results of varying heating rate ose experiments
on PET, Nealy et aI. 11 reported that sorne form of restructuring of the higher melting
material was occurring during the thermal analysis scan. This suggestion was derived
from the observation of a dependence of the lower temperature peak position and the
crystallite size on the initial crystalIization conditions which was construed as a
demonstration of sorne continuity of initial structural feature upon the conversion of the
higher melting material to the lower melting material.
It was Ikeda et aI.4 on the basis of DSe studies on PET, who fust postulated that
the higher temperature endotherm represented the melting of a structure not present
initially in the sample, but formed through annealing or recrystallization of the partially
melted crystalIites during the programmed heating scan. The lower temperature
endotherm, therefore, represented the relatively disordered folded-chain material initially
present in the sample. This study was also unique in its discussion of the effeet of the
instrument heating rate on the kinetic balance between melting and recrystallization of
partially melted regions, pointing out that the observed compound endotherm was the
algebraic sum of these simultaneous processes.
3-5
•
Thermal Rt'llIIl'ior
3.1.1.3 Reorganization Madel
The cont1icùng raùonales of the multiple melting phenomenon in lhe SIII/II' pnlymer
system distinctly showed lhe neccssity for syslematic conlrol of lhe variables. naIncl)'.
cryslallizing, annealing, and heating condiùons. In 1970, Robens l2 presenled lhe tirsl
clear demonstration of the anncal-type rearrangemenL~ taking place in lhe PET, during the
thermal analysis heating scan. A simple experiment designed 10 explore spccilically the
processes that transpire during the ose heating scan was pert·ormed. A distincl exotherm
was displayed in the thermogram when the heating scan of isothermally crystalli7,cd PET
was interrupted and the temperature held constant between the two endothermic peak
temperatures. The results of varying heating rate experimenL~ demonslraled thal wilh
increasing heating rate, a proportionately smaller amounl of crystalliles ha.~ ùme 10
recrystallize during the ose scan, thus accounting for the accompanying increa.~e in
magnitude of the lower temperature endotherm. Holdsworth and Turner-Joncs 13 also
explored the in-scan conditions of isothermally crystallized PET with simultaneous oseand X-ray studies. The X-ray data complimented the original findings of RoberL~12 hy
providing evidence of continuously increasing overall crystallite perfection during heaùng
of the sample in the ose over the temperature range between the two peaks.
From the ose studies of PET using varying heating rates, Sweel and Belll~
reported that in general the rate of melting was determined by the heating raie, while lhe
rate of recrystallization was dependent on the degree of undercooling. An ultimate
recrystallization temperature was defmed as the temperature just below the higher
temperature peak, where the rate of melting approaches the rate of recrystallization. The
rate of meIting finallyexceeds it at the higher melting peak temperature.
The above studies renounced the earlier hypothesis of the conversion of
kinetically-favored higb-melting material to thermodynamically-favored lower-melting
material for the case of PET, and other structurally different polymers. More importantly,
3-6
•171erma/ Behavior
these studies pointed out that the DSC thermograms are not necessarily a direct rellection
of the state of the material at room temperature prior to thermal analysis.
The focus of the relatively new research efforts is to explore the mechanisms of the
structural rearrangements of the crystallites. Supplementing the DSC heating thermogram
information with data from other methods of analysis permil~ the refinement of the
understanding of concomitant melùng and recrystallization. For example, the overall
crystallinity of the sample can be determined by WAXS, while the smaller scale
informaùon provided by the method of smalI-angle X-ray scatlering (SAXS) a1lows the
determination of lamellar thickness. The feasibility of performing simultaneous SAXS and
DSC techniques has been demonstrated by Russell and Koberstein on polyethylene. 15 The
combination of DSC, WAXS and SAXS analyses have proved useful in the elucidation of
the multiple melting behavior exhibited in poly(ether ester ketone) (PEEK),22 whose
melting behavior is very similar to PET.13 In a typical DSC experiment, the heating of
isothermally crystallized PEEK samples was halted at various temperatures between the
two melting peaks, and the samples were immediately quench-cooled to the starting
temperature. The heating scan was subsequently repeated to attain the DSC scan
representative of the sample present at the original hait temperature. A series of haIt
temperatures were employed in an effort to obtain a profIle of the sample as it was being
heated. It was found that the low temperature peak accordingly marked the point at
which the original isothermally formed crystallites became unstable and where the onset of
melting and recrystallization occurred. Both the crystallÎlÙty and the lamellar thickness
were determined to increase systematically with increasing temperature.
Evidently, PET has been the polymer of historical importance in the development
of models of multiple melting in bulk polymers. The model of reorganization of
crystallites during the thermal analysis scan has becn verified as the origin of the similar
multiple melting behavior exhibited in i-polystyrene,17.19 and in PolY(E-caprolactone) in an
elegant analysis presented by Rim and Runt.21 The same model also fits weil with the
3-7
•
•
•
Thenlla/ 8ellllvior
observed thennal behavior of poly(butylene terephthalate).22 However. the
recrystallization model cannot account for all multiple melting behavior. Different ~'r)'slul
structures have been ascribed as the source of the double melting behavior of rrans-I,4
polyisoprene23 and poly(vinylidene fluoride).24 Chung and Cebe described the multiple
melting behavior in poly(phenylene sulfide) by including effects of both morphology and
reorganization.25-26
In a more recent work by Rodriguez-Arnold, et al.27 il was discovered that in
isothermally crystallized syndiotactic polypropylene, reorganization of the crystalline
material takes place during the DSC scan. However, this transfonnation involves chain
motion to achieve perfection at the level of the Imit cell. ln a remarkable process
involving changes to both the surface and the internai structure of the crystal\ile, the
pseudo-chiral chains alter their packing to achieve the more stable antichiral packing, in a
type of solid diffusion process driven by the lower energy configuration of the lutter
arrangement
In an elegant study by Caldas et al.28, the characteristic double melting endothenn
behavior of quench-cooled isotactic polypropylene (i-PP) was investigated using DSC and
solid state 13C nuclear magnetic resonance spectroscopy (NMR). Heating a sample
subsequent to armealing it at a temperature corresponding to the lower ternperature
endotherm maximum yielded a thermogram with a single melting endothenn. A variable
temperature solid state NMR investigation indicated dramatic changes among the
crystalline phase spectra of samples studied at temperatures corresponding to those
between the two endotherm maxima observed in the DSe. The double endotherm
behavior of i-PP was attributed to the presence of the two crystal forms (al and ~) and
their interconversion during annealing.
3-8
•Thermal Behavior
3.1.2 The Present Study
Prcscntcd in this chapter is a comprehensive look at the multiple mdting hehavior
routinely ohserved in samplcs of the isothermally melt-crystallized poly(epichlorohydrin)
(PECH) polymers, using differcntial scanning calorimetry (DSC). The dependence of the
ohscrvcd mdting behavior of the polymers on the experimental conditions employed i:l
this study was found to be rather complex. It is the aim of this work to explore the
metastahility of the crystallites in order to leam more about any structural distinctions
among the differcnt forms of the PECH polymers. Varying heating rates, crystallization
times, and undercoolings are used to explore the multiple melting hehavior observed in the
optically pure polyenantiomers poly(R-epichlorohydrin) (PRECH), and poly(S
epichlorohydrin) (PSECH); their equimolar blend; and the stereoblock poly(R,S
epichlorohydrin) (i-PRSECH). A brief description of the thermal behavior of the cold
crystallized PECH polymers is also included. The thermal behavior of the chemicall~
derivcd poly(propylene oxide} (PPrO) is treated separately, and the effect of the removal
of the chlorlne atom on the polymer thermal behavior is discussed.
3.2 Experimental
The details of the differential scanning calorimeter instrument and the calibration
standards employed in the investigations of the thermal behavior of the polymers have
bcen presented in Section 2.3.6. The information pertaining to scanning rates and
temperature programs is specific to a thermogram in question; therefore it is reported in
the text and/or the figure captions in the following Section. In ail cases, the positions of
the peaks in the thermograms are reported as the temperature at the peak maximum.
3-9
•Thermal Beha\'Ïor
3.3 Results
3.3.1 Poly(epichlorohydrin) Cold-Crystallization
Figure 3.I(a) contains the DSC thennogram for the cooling of the PSECH
polyenanùomer at a rate of 20 oC/min from the melL The thennogram shows nu sign IIf
crystallizaùon exothenn and essenùally demonstrates the ease with which this polyrner can
he quench-cooled from the me1t to fonn an amorphous glass. Vpon quench-cooling at a
nominal rate of 150 oC/min from 175 oC to -50 oC, ail of the PECH polyrners are
essentially amorphous. Figures 3.l(b) and (c) contain the representative DSC heating
thennograms for the PSECH polyenantiomer and the stereoblock i-PRSECH, respcctively,
obtained during subsequent heating at a scanning rate of 20 oC/min. The cOITesponding
thennogram of the PRECH polyenantiomer is the same as that shown for PSECH. The
PSECH thennogram displays a single crystallization exothenn at approximately 25 oC and
on continued heating reveals a single melting endothenn at 112 "C. The equimolar
polyenantiomer blend has the same transition temperatures, with a slightly broader
crystallization exothenn. For i-PRSECH, the thennogram is similar to those for the
polyenantiomers, but the corresponding exothenn and endothenn values are shifted to
approximately 50 oC and 106 oC, respectively. In ail of the OSC traces, the crystallization
exothenn and the melting endothenn are equal in magnitude, verilying that no
crystallization occurred during quench-cooling.
3.3.2 Multiple Melting in Poly(epichlorohydrin)
3.3.2.1 First Melt Thermograms
The isotactic poly(epichlorohydrin) polyrners crystallize out of solution during
polymerization. Figure 3.2 contains the fJrst-melt OSC thennograms of the as
polymerized PRECH and PSECH, and of the isolated i-PRSECH fraction. At a heating
3·10
•Thermal Behavior
(a) -
-
-(b)
o
~
":25 o 25 50 75 100 125 150
Temperature (OC)
Figure 3.1 DSC cooling thennogram of (a) PSECH (scanning rate = 20 oC/min),
(b) heating thennograms of PSECH, and (c) i-PRSECH both quench
cooled from the melt (heating rates =10 oC/min).
3-11
•
r(a)--------(b)
(c)
(d)
Thermal Behavior
25 50 75 100 125 150
Temperature (OC)
Figure 3.2 DSC flI'st-melt Ùlermograms of (a) Ùle as-fractionated i-PRSECH, (b) the
as-polymerized PRECH, and (c) Ùle as-polymerized PSECH. ail heated at a
rate of 20 oC/min. In (d) Ùle as-polymerized PSECH is heated at a rate of
5 oC/min.
3-12
Thermal Behallior
rate of 20 oC/min all of the PECH polymers display two peaks in the melting endotherm.
The DSC trace of the i-PRSECH in (a) is distinct from the thermograms for the
polyenantiomers in (b) and (c) with a dominant lower temperature peak, and an overall
broader endotherm. In addition, the higher temperature peak of i-PRSECH falls at
117 oC, approximately 4 oC higher than that of the polyenantiomers. The higher melting
temperature is most likely a result of fractionation during separation from the atactic
materiai, a procedure which was not performed with the polyenantiomers. Figure 3.2(d)
demonstrates the effeet of a slower heating rate on the appearance of the ftrst-melt
thermogram of PSECH. At a scanning rate of 5 oC/min, the sample displays a single peak
endotherm at a slighùy higher temperature and has an indication of a shoulder on the low
temperature side.
3.3,2.2 Varying Heating Rate Experiments
The effeet of varying the heating rate on the observed melting behavior was
determined for PSECH and the equimolar polyenantiomer blend. Samples isothermaily
crystallized at 90 oC for 60 min were heated at rates ranging from 1.2 to 40 oC/min. The
resultant thermograms for PSECH and the equimolar polyenantiomer blend are presented
in Figures 3.3(a) and (b) respectively. At heating rates of 10 oC/min and lower, the
polymers exhibit triple melting peak endothenns. The flfst peak is indicated in the
thennogram with an arrow, and tends to be somewhat obscure at low heating rates. The
position of this peak is relatively insensitive to the change in heating rate, and will be
considered later. Therefore, the terms lower and higher temperature endotherm are used
in reference to the area under the seeond and third peaks, respectively, of the multiple
melting endotherm.
For both of the samples, the area of the lower temperature endotherm grows with
an increase in heating rate. The effeet of varying the heating rate on the positions of the
peak temperatures for PSECH is demonstrated graplùcal1y in Figure 3.4. The positions of
3-13
• (a)
40 oC/min
1 10 oC/min
0"'0
~5.0 oC/min
e
1.20C1nùn
(b)
1.2 oC/min
77rerma/ Behal';or
90 100 110 90 100 110 120
Temperature (OC)
Figure 3.3 DSC thennograms of (a) PSECH and (b) the equimolar polyenantiomer
blend. isothennally crystallized at 90 oC for 1 h and heated at varying rates.
3-14
•Thermal Behavior
120..,......------------ï
•115,......,Ua'-"
~:::3 110~"""euc..e~
105
'.
...............................0......:::·:0..······.------....... .
o 10 20 30 40
Heating Rate (OC/min)
Figure 3.4 The lower (e) and lùgher (.) peak temperatures as a function of DSe
heating rate for PSECH isothermally crystallized at 90 °e for 1 h. Hollowsymbols = single peak temperature.
3-15
•Thermal Behm'ior
the two endothennic peaks shift in opposite directions with increasing heating rate. With
increasing heating rate. the position of the higher temperature peak shifL~ to lower
temperatures faster than the corresponding shift to higher temperatures of the lower
temperature peak. Both the lower and the higher peak temperatures depend linearly on
the heaùng rate and the data can be extrapolated to predict a single peak mdting
endothenn at a heating rate of about 20 oC/min. The experimentally ohserved tempemture
of the single peak obtained at a heaùng rate of 40 o(,Jmin falls on the extmpolation of the
line fitting the lower temperature peak data.
Table 3.1 contains the heats of fusion of PRECH. the equimolar polyenantiomer
blend, and i-PRSECH saIrples. all isothennally crystallized at 90 oC for different periods
of time and subsequentjy heated at a rate of 20 OC/min. Under these conditions, the
thermograms of each of the polymers display single melùng endothenns (as shown for the
20 oC/min thennogram in Figure 3.3), with the peak position independent of
crystallization time. The rate of development of crystallinity in the polyenantiomer
exceeds that of the equimolar polyenantiomer blend. Similar heats of fusion are indicated
at a crystallization lime of at least 60 min. By contrast. the heat of fusion attained by the
stereoblock polymer is only a fraction of the levels attained by the other polymers. Afler a
crystallization period of 20-22 h. the heat of fusion attained by the stereoblock polymer i~
only about 85 % of that attained by the other polymers.
Table 3.1 Beats of Fusion of PECH Crystallized at 90 oC for Different Times
heat of fusion. !:JI (J/g)i-PRSECH 50/50 Blend PSECH
trace 1.6 3.6trace 2.3 5.6trace 9.7 n.d.1.4 30 3211 37 3937 45 43
crystallization lime
15 min20 min
30 min1h2h
20-22h
3·16
•
•
Thermal Behavior
3.3.2.3 Partial Heating Experiments
In an attempt to obtain a melting protile which is representative of what is
occurring as the samplc is hcatcd in the OSC. the following parùal heating experiments
wcre pcrformed. A sampie of PSECH was isothermally crysta11ized at 90 oC for 60 min in
the OSC, and subsequently heated at a scanning rate of 5 oC/min [Figure 3.5(a)]. The
higher temperature peak appears at about 112 oC. As Table 3.1 indicates, the isothermally
crystallized material in the parùal heaùng experiments (Tc = 90 oC for 1 h) does not
represent a completely crysta11ized sample. The OSC traces in Figure 3.5 are presented in
a shifted fashion but are all on the same scale. The thermogram shown in Figure 3.5(b) is
a hcaùng scan at 5 oC/min, as in (a), but stopped at 112 oC. corresponding to the higher
temperature melùng peak. The sample was then immediately quench-cooled to the
original crystallization temperature of 90 oC. The OSC was allowed to equilibrate for
1-2 min before repeating and compleùng the heaùng scan. The resulting thermogram•
presented in Figure 3.5(c), shows a broad endothenn preceding the stop temperature,
followed by a sharp higher temperature peak at 114 oC. Fmally, the heating scan at
5 oC/min was repeated with another sample and stopped at the same higher temperature
(112 OC), but held isothennal for 56 min before quench-cooling the sample to the original
crystallization temperature. The OSC was allowed to equilibrate for 1-2 min before
repeating and compleùng the heaùng scan. The resulùng thennogram. presented in Figure
3.5(d), shows a small, broad peak corresponding to the 10wer temperature. The higher
temperature endothenn exhibits a remarkable narrowing in shape, and the peak position
has increased to abave 116 oC. The area under the higher melùng peak in Figure 3.5(d) is
greater than that in Figure 3.5(a), indicating that a longer exposure to the anneal
conditions at 112 oC results in a greater amount of recrystallization of the sample.
In a separate experiment, a sample of PSECH was isothermally crysta11ized at
90 oC for 60 min in the OSC, and the subsequent 5 oC/min heating scan was stopped at
the temperature of the observed minimum between the two endotherms, Le. at 112 oC.
3-17
17lermal Behavior
,1
1" 1'.'
(b)
(a)
100 110 120
Temperature (oC)
Figure 3.5 Dse thennograms of PRECH isothennally crystallized at 90 oC for 1 h and(a) heated at 5 oC/min, (b) heated at 5 oC/min until high temperature peak,
(c) same as (b) and immediately quench-cooled to 90 oC and subsequently
heated at 5 oc/min, and (d) same as (b) and held at the high temperature
peak for 56 min, quench-cooled to 90 oC, and heated at 5 oC/min.
3·18
•Thermal Behavior
The sample was held isothermally for 42 min before resuming the heating scan of the
sample. The heating and holding OSC thermogram is presented in Figure 3.6 [The
uninterrupted scan is prescnted in Figure 3.3(a)]. Ouring the isothermal hold an exotherm
is visible, indicating the presence of recrystallization events in the temperature region of
the observed minimum in the melting endotherm. Completion of melting of the sampIe
resulL~ in a double melting peak endotherm.
3.3.2.4 Varying Crystallization Temperature Experiments
The effect of varying the crystallization temperature (undercooling) on the shape
and position of the melting endotherms of PRECH, the equimolar polyenantiomer blend,
and i-PRSECH is demonstrated with the collection of OSC thermograrns in Figure 3.7.
For the purposes of presentation, the y-scale of the thermograrns has been adjusted to
make ail of the endotherms similar in amplitude. The samples were isothermally
crystallized for ca. 24 h at eight different undercoolings ranging l'rom 48 oC to 20 oC
(thermograrns at selected undercoolings are shown here for each sample). The samples
were subsequently heated at a scanning rate of 20 oC/min. Typically, the i·PRSECH
samples were of much lower crystallinity than the polyenantiomers or the equimolar
polyenantiomer blend. At an undercooling of 28 oC (Tc = 110 OC), the i·PRSECH
stereobloek does not crystallize to any significant degree after 24 h. Therefore, the
smallest undercooling employed for this polymer was 33 oC (Tc = 105 oC). It is important
to compare the thermograrns of the samples erysta1lized at 90 oC (heated at 20 oC/min) in
Figure 3.7 to the corresponding thermograms heated at a rate of 20 oC/min (Tc = 90 OC) in
Figure 3.3. It is noted that the insignijicant fust peak in the thermograrns in Figure 3.3
corresponds 10 the lower temperature peak in the double melting peak endotherms in
Figure 3.7. In Figure 3.7, al a given undercooling, the positions of the lower and higher
peak temperatures are the same for ail of the polymers, but the ratio of the areas of the
lower and higher temperature peaks differs. In ail of the thermograms, the shape of the
3-19
•Thermal Behavior
IV \.---1
• 1 1
90 100Temperature (oC)
110 120Temperature (oC)---------
1 i i i i 1 i i i i i i i i , 1 i i i i 1 i i i i 1
o 5 10 15 20 25
"fiIœ (nùn)
Figure 3.6 DSC thennogram of PRECH isothennally crystallized al 90 oC for 1 h and
heated al a 5 oC/min until 110 oC, he1d isothennal 42 min. and then
resumed heating al 5 oC/min.
3-20
t
Thermal Behavior
{a) ,J (b) {'v (c)
,-\....r \.....-
rf-..~no oC 110°C
105 oC\105 oC,
~\105 oC,-
~
r\ f\,1
\ 100 oC
~\ 100 oC rJ, \100 oC
VJ \ 90°C \ 90°C J \ 90°C
1 1 1 1 1 1 1 1 1
100 110 120 100 110 120 100 110 120 130Temperature (OC)
Figure 3.7 DSC thennograms of (a) PRECH, (b) the equimolar polyenantiomer blend,
and (c) i-PRSECH samples isothennally crystallized at Tc for ca. 24 h andsubsequently heated at 20 oC/min.
3-21
•Thermal Behlll'ior
double melting peak thermogram changes with incrcasing crystallization temperature: the
lower temperature peak becomes progressively dominant. The ratio of the lower to higher
temperalUre peak is consistently higher in i-PRSECH than in the olher polymers. The
effect of varying the undercooling on the melting protiles of the cquimolar polyenantiomer
blend and PRECH is essentially the same, until the smallest undercooling is employcd. At
the smallest undercooling of 28 oC (Tc =110 OC), the lower lemperature endotherm is
very slightly more dominant in the thermogram for the hlend, whercas in the PRECH
thermogram the higher temperature endotherm is the dominant peak. Far less suhtle is the
distinction belWeen the melting profùes of i-PRSECH and PRECH or the cquimolar
polyenantiomer blend. In the i-PRSECH thermogram, the lower temperature endotherm is
prominent at undercoolings as great as 38 oC (Tc = 100 oC).
The effect of varying the undercooling on the position of the peak temperatures in
Figure 3.7 has already been demonstrated graphically in Figure 2.14, the Hollinan-Weeks
plot for the PECH polymers. The lower temperalUre peak depcnds linearly on the
crystallization temperature and the slope is roughly parallel with the Till = T, line, wherc Till
is the observed melting temperature and Tc is the crystallization lemperature. On the other
hand, the position of the higher temperature peak increases less rapidly with crystalli7.ation
temperature and consequently has a slope of less than 1.
The changes resulting from varying the crystallization temperature can also rcsult
from thermally annealing the sample at different temperatures.
3.3.2.5 Poly(propylene oxide) (PPrO)
The Dse thermogram for poly(S-propylene oxide) (PSPrO) cooled from the melt
at a rate of 20 oC/min is shown in Figure 3.8(a). A crystallization exotherm is observed at
about 5 oC. Although not shown here, on subsequent heating at a rate of 20 oC/min, the
DSC trace shows only a small exothermic deviation from the baseline at the onset of a
3-22
Thermal Behavior
o"0
&5
(a)-~ Î{
- (b)
- (c)
- (d)
• 1 l ' 'l' • l'" 1 •• 1 1
-50 -25 o 25 50 75 100
Temperature (OC)
125 150
Figure 3.8 DSe thermograms of (a) PRPrO cooled from the mell al 20 oC/min, (b)
PRPrO, (c) PSPrO. and (d) i-PRSPrO quench-cooled from the mell and
heated al a rate of 10°C/min.
3·23
•Tlrermal Belral';or
single peak melting endotherm with a maximum at 63 oC. The areas of the exothermic
transition on cooling and of the endothermic transition on heating are approximatdy cqual.
Figures 3.8(b), (c), and (d) contain the heating thermograms of PSPrO, PRPrO, and
i-PRSPrO, respectively, recorded at a scanning rate of 20 oC/min, al'ter cooling the meh
l'rom 150 oC to -50 oC at a nominal rate of 150 oC/min. An exothermic deviation l'rom the
baseline is apparent at the onset to the relatively large single peak endotherm which has a
maximum at 63 oC for ail of the polymers.
In general, il is found that the rates of crystallization are much greater in the PPrO
polymers as compared with the corresponding PECH polymers. Typically, the
isothermally crysta1lized PPrO polymers yield single peak melting endotherms under
normal DSC scanning conditions. However, isothermal crystallization of PPrO for
relatively short periods of time (ca. S 20 min) gives rise to samples that exhibit multiple
peaks in the melting endotherm upon subsequent heating. Figure 3.9 is an example of the
multiple melting behavior found to occur in POO. In this figure, the i-PRSPrO polymer
samples were isothermally crysta1lized from the melt for 20 min at temperatures mnging
from 20 to 45 oC, corresponding to undercoolings of 62 to 37 oC, and then heated at a
scanning rate of 10 oC/min. At crystal1ization temperatures of less than 40 oC, a small
endotherm is apparent just above the crystal1ization temperature. The significance of this
peak is discussed later. The re.~ponse of the other two peaks in the thermograms to the
change in Tc is typical of annea1-type behavior. The higher temperature peak remains
relatively unchanged with increasing crystal1ization temperature, while the position of the
lower temperature peak is increased in temperature. At the highest crystallization
temperature employed, the lower temperature peak is predominant.
3-24
•
•
r
Thermal Behal'ior
30°C
25°C
20 30 40 50 60 70 80
•
Temperature (OC)
Figure 3.9 DSC thennograms of i-PRSPrO isotherma1ly crystallized at Tc for 20 minand heated at a rate of 10°C/min.
3-25
•
•
•
Thermal Bellav/or
3.4 Discussion
3.4.1 Effect of Varying the Heating Rate
In Figure 3.3, the opposite shift directions observed for the lower and higher
temperature peaks accompanying the increase in heating rate, conform to previous
findings of multiple melting investigations of other polymers.7.14.17.23 The change in
position of each of the peaks is discussed separately. The origin of the downward shift of
the higher temperature peak with increasing heating rate can be expl:ùned as follows:
With increasing heating rates, the sample passes through the annealing temperature region
so quickly that there is little opportunity for reorganization of the material originally
formed at Tc' Accordingly, the extent of reorganizatioll decreases so that the high
temperature melting material is less ordered and melts at a lower temperature.
The increase in the lower temperature peak with increasing heating rates has been
explained in two ways. According to both rationales, the lower temperature peak is
assigned as that temperature where the original isothermally-formed crystallites become
unstable, and to the onset of (partial) melting and recrystallization.16.21.29 (1) In one
explanation, the increase in temperature of the lower temperature peak is rationalized
entirely in terms of superheating: Owing to their large surface-to-volume ratio. and hence
large number of defccts at the surface, the original small imperfcct crystallites show a
reduced entropy of fusion on melting.29 With increasing heating rate, more of the original
imperfcct material remains, and this material is presumed inherenùy susceptible to the
problems of thermal conductivity. While this rationale is easily accepted in its intuitive
nature, it is perhaps an oversimplification of a situation more accurately described by
others in the following. (2) Rim and Runt21 have developed the interpretation below:
Their explanation is self-consistent with the heating rate dependence of the shape and
magnitude of bath endothermic peaks. and is shown schematically in Figure 3.10. The
mechanism depicts both the shifts in the lower and higher temperature peaks as
3·26
•Thermal Behavior
Heating Rate
Slow Intennediate Fast
~/ ...··\.;vtr Mr
,: AI .... :A···""".: ,..\ Al ;' -"':....:-..•A••,.....-.__~•• "-=.:.-
'. :.......C C· .· .· ....
No annealing
M = melting of original crystalsMr =remeltingC =recrystallization
Tille> Tmb > Tilla
Figure 3.10 Schematic illustration of the Rim and Runt21 model of reorganization in
polymers during the thermal analysis heating scan.
3-27
•Thermal Behm'ior
consequences of a complex. composite melting endotherm. The ohserved endotherm i~
the net illustraùve effect of three thermal processes: the melting of the original imperfcct
crysta11ites (M). the recrystallizaùon of sorne or ail of this material (C). and linally the
melùng (Mr) of the recrystallized material. The lower mc1ting endOlherm rcprcsents the
melùnglparùal melting of the iniùal metastable crystallites. Whercas the size of this tirst
endotherm is relaùvely independent of heating rate. the magnitude of the following
exotherm however. is nol. The magnitude of the recrystallization exotherm (AI at slow;
A2 at intermediate heaùng rates) determines the magnitude of the final endotherm. At
slow heaùng rates. there is sufficient lime during the DSC scan for the polymer to
reorganize and consequently the net thermogram displays a large higher temperature peak.
Under these scanning conditions the lower temperature peak may appear as a shoulder on
the low temperature side. At intermediate heaùng rates. there is less Ùme for the
reorganization processes to occur and the thermogram will depict a smaller higher
temperature peak. Fmally. it follows that at fast heaùng rates. the reorgani7.ation process
is restricted, minùnizing the recrystallizaùon exotherm and thus the melting endotherm of
the recrystallized material. The resultant thermogram displays a large lower temperalUre
peak. This rationale takes into ac;:ount the limits of DSC, i.e., it is recognized that the
observed thermogram is not only a function of the thermal processes ongoing in lhe
sample, but also an algebraic SUffi of the endo- and exothermic transitions. In general, the
net effect of the recrystallization exotherm is to shift the flrst melting peak to lower
temperatures as a consequence of competing simultaneous processes. Consequently, the
slower the heaùng rate, the larger the recrystallization exotherm, and the more down
shifted the tirst melting peak, i.e., Tme > T"" > T....
As stated in the introduction, the original polymer sample usually includes a
distribution of crysta11ite perfection. To this end, a broad melùng endotherm would he
observed upon heaùng, even in the absence of reorganizaùon during the scan. The
process of reorganization during heating will anneal all or part of the original material into
3-28
•Thermal Behavior
a narrower distribution of crystaIlite perfection, which subsequently melts at a higher
temperature. Although the model in Figure 3.10 demonstrates the full separation of the
melting peaks corresponding to the original and recrystaIlized material, il is possible that a
portion of the initial distribution of crystaIlites will melt at the same temperature as the
rccrystallized crystallites.
In terrns of the results presented in Figures 3.3 and 3.4, the shift to lower
temperatures of the higher peak with increasing heating rate is readily accounted for by the
above explanation. The change in position of the lower temperature peak is probably a
combination of both explanations presented above, with thermal conductivity considered a
factor only at intermediate and higher heating rates.
A possible criticism of the superheating mechanism presented earlier is the tacit
dismissal of this phenomenon becoming a factor in the position of the higher temperature
peak as the heating rate is increased. It is understood that the material represented by the
higher temperature endotherm hecomes increasingly \ike the original imperfect material
with increasing heating rate. Thus, it should he sensitive to the same problems of thermal
conductivity, namely superheating. In essence, superheating of the higher temperature
peak material would counteract the observed decrease in peak temperature. If this is
indeed the case, then the question must he asked, why the higher temperature peak is
observed experimentaIly to he more sensitive to the change in heating rate. Naturally, at
excessively fast heating rates, thermal conductivity of the intrinsically insulating polymer
sample will most \ikely play a role and the elevation of the single peak endotherm in the
40 oC/min scan in Figure 3.4 is probably an artifact of superheating.
3.4.2 Effect of Partial Heating
A technique of partial heating was originally practiced by Lemstra17 in an attempt
to provide incontrovertible evidence of recrystallization during heating in the thermal
analysis of i-PS. Typically, the sample is heated to a temperature just above the lower
3-29
Thermal Belravior
temperature peak, quench-cooled, and reheated to complete mdting. Accorùing III
Lemstra and coworkers, if the double melting hehavior originates l'rom the mdting of two
distinct crystal systems, then only the high temperature peak is expected III rcmain upon
subsequent reheating. More importantly, il will possess the same heat of fusion as in the
original thermogram. For the PSECH sample in Figure 3.5, the magnitude of the higher
temperature peak increases upon reheating in (c). The presence of a broad endotherm
helow the original stop temperature indicates that not ail of the material which had melted
had also recrystallized by that point in the original heating scan, and probably
recrystallized upon quenching. If the sample is held at the stop temperature für an
adequate period of lime to permit the partially h1~lted material to rcorganize into the
material represented by the higher temperature peak, the thermogram rcsulting l'rom
subsequent heating should contain much less of the broad, lower temperature endotherm.
This is demonstrated experimentaily in Figure 3.5(d). This test has heen used in support
of the recrystailization model in studies of PEEK29 where it was demonstrated that the
lower temperature endotherm represented only a portion of the melting of the original
crystallites that exist in the sample prior to scanning.
It might he argued that the high temperature peak is simply representative of the
melting of material isothermally crystallized at the hold temperature, since the original
sample was not completely crystailized. However, the area under the high temperature
peak [Figure 3.5(d)] is an order of magnitude larger than the:: area under the full melting
endotherm of a sample isothermally crystallized at approximately the holding temperature
[Tc = 110°C in Figure 3.7(a)], for 24 h. Secondly, the position of the high temperature
peak in Figure 3.5(d) is lower than that of the isothermally crystallized sample in Figure
3.7(a), indicating that the lamellae of the former have not thickened to the extent of the
lamellae formed in the isothermally crystallized sample. Clearly, the thermogram in Figure
3.5(d) is representative of a recrystailization process which can occur during the scan, and
3-30
•Thermal Behavior
which can occur to a greater extent if the sampIe is held isothennally for a suflicient time
period, as in Figure 3.5(d).
In Figure 3.6, the observation of an exothenn between the two endothennic peak
temperaturcs supports the proposai of recrystallization events transpiring during the DSC
scan. Similar exothenns have been reported in relatively few cases. 12 Once again, it is
important to note that it would take more than a week to completely isothennally
crystallize the sample at the holding temperature of 110°C. Therefore, the area of the
endothenn following the isothennal hold in Figure 3.5 is related to a recrystal1ization
proccss and not simply to the primary crystal1ization of the original sample. It is very
difficult to support the IWO distinct crystal lattice systems model for the multiple melting
behavior of PECH in Iight of the results of Figures 3.3 to 3.6.
3.4.3 Effect of Varying Crystallization Temperature
The results of the varying heating rate experiments (Figure 3.3) and the varying
crystal1ization temperature experiments (Figure 3.7) each demonstrate the double melting
endothenn behavior, but in different temperature regions of the therrnogram. Specitically,
in Figure 3.3, it is shown that at relatively large undercoolings, a triple peak endothenn is
observed upon subsequent melting of the sample, and the shape is strongly dependent on
heating rate. The highest temperature peak fonns from a continuous recrystallization of
the middle temperature peak material during the heating scan. In Figure 3.7, it is shown
that at relatively small undercoolings, a double melting endotherrn is observed upon
subsequent melting of the sample, and the shape is dependent on crystallization
temperature and crystal1ization time. Dnder these conditions the higher temperature
endothenn does not recrystallize during the scan.
AlI of the double me1ting endothenns in Figure 3.7 were collected using a heating
rate of 20 oC/min. It bas already been demonstrated by the results in Figure 3.3, for
PSECH and the equimolar polyenantiomer blend, that a beating rate of 20 oC/min does not
3-31
•Thermal Behav;or
pennit an isothennally crystallized (Tc =90 OC) sample suflicic:nl exposun: Ume 10 lhe
anneal conditions to pennit recrystallization during the scan. During suhsequent heating
of the samples which have becn isothennally crysl;illized al smaller undercoolings in
Figure 3.7. the drive to recrystallize upon heating is even less. For example. when lhe
thennogram of the equimolar polyenantiomer blend sample isothermally crystallized al
110 oC and heated at 20 oC/min is compared. in Figure 3.11. with the thennogram of the
same sample recorded with a heating rate of 5 oC/min. no difference in the ratio of lhe
lower and higher temperature peaks is detected. indicating no reorganization. The
question thus arises as to the origin of the double melting peaks in the thennogmms of
Figure 3.7. According to the above. the double melting peak endothenn is delennined hy
the distribution of crystallites present in the sample before thennal analysis. Therefore.
during isothennal crystallization. at least two distinct levels of crystallite perfection are
fonned.
3.4.3.1 Secondary Crystallization
The role of secondary crystallization is discussed in the context of the
thennograms of the isothennally crystallized sarnples shown in Figure 3.7. Secondary
crystallization has been defmed30,31 as a slow crystallization process occurring towards the
end of the main (prlmary) process. involving the reorganization of the crystallites into
larger and more perfeet ones. In the present case, if seeondary crystallization involves"a
process of perfection of the lower temperature peak to the higher temperature peak. the
observed position of the latter should increase accordingly with isothennal crystallization
lime. This does not occur. Rather, as Figure 3.12(a) demonstrates for sarnples of PRECH
crystallized at 90 oC for a lime period ranging l'rom 5 min to 20 h, the position of the
lower temperature peak increases slightly at tirst, before remaining constant, and the
higher temperature peak remains unchanged.
The tenn seeondary crystallization is also used to detine an opposite
3-32
•Thermal Behavior
(a)
115 120 125 130
Temperature (OC)
Figure 3.11 ose heating thennogram of the equimolar blend sample isothennally
crystallized at 110°C for 24 h and heated at (a) 20 oC/min and
(b) 5 oC/min. (both thennograms calibrated with indium at 20 oC/min.)
3-33
•Thermal Behlll'ior
t0
120 min"t:l
~ 60 min
30 min
• 27 min25 min-- 20 min
15 min
'-....5 min
2 min
100 110 120 130
Temperature (OC)
Figure 3.12 (a) OSC thennogram of PRECH isothennally crystallized at 90 oC l'or
different periods of lime and heated at 20 oC/min and (b) plot 01' partial
heat of fusion of the lower and higher temperature peaks in (a) as a
function of crystallization lime.
3·34
Thermal Behavior
• 50........01)
;:::; 40'-' •s:::0 30 .,.-CIl •~ higherpeak
..... 200..... /lowerpeak~ 10 •Q) 0::z:: • 0 0
00 50 100 150 1000 1100 1200 1300
Time (min)
Figure 3.12 ...continued.
mechanism.28.29 During isothennal crystallization. dominant lamellae fonn initially.
Secondary crystallization of initially rejecte<! material follows. with the development of
subsidiary lamellae which. through branching. fiII the spaces between the dominant
lamellae. The data presente<! in Figure 3.12 (b) are in accordance with this interpretation.
The areas of the lower and higher endotherms are plotted as a function of crystallization
time. for the PRECH thermograms in Figure 3.12 (a). It can be seen that the higher
endothenn is representative of material which develops before and at a faster rate than that
represented by the lower temperature endothenn.
In their study of i-PS. Pelzbauer and Manley19 found a similar dependence of the
specific heat of fusion on crystallization lime for the multiple peaks. The latent and slow
rate of crystallization of the tirst peak was described as being characteristic of impurities
3-35
Thermal Beh{/vior
rejected during prirnary crystallization. However. the peak is slill present in lhe melting
thermogram of a sample crystallized from dilute solution 17• Similar data for i-PS have
been reported by Lemstr.l et al. IH • The two levels of perfection. represenled hy lhe tirsl
and second melting peaks in the thermograms of the samples crystallized at relatively lnw
undercoolings. were attributed to the presence of suhsidiary and dominanl (mnst perli:ct)
lamellae. respectively. This interpretation ims mosl recently been used to explain the
similar effects of varying the crystallization temperature on the multiple melting behavior
in poly(ethylene succinate).32 The rational was substantiated by evidence from scanning
electron microscopy of the isothermally crystallized samples: a highly brancht..'d spherulitic
structure developed in the samples only after crystallizing fer a period of time al kast
equal to the induction lime for the appearance of the lower lemperature endotherm in the
DSe. In a contemporary study. computer simulated DSe heating thermograms
successfully predicled the appearance of both the triple and the double melting peak
hehavior experirnen!ally observed in poly(butylene terephthalate).33 The model was based
on a heating rate recrystallization dependence of an initial distribution of melting
temperatures of a single crystal-type material. In the triple melting peak thermogram. the
third (highest temperature) peak was judged to he a representative of the melting of
material which had recrystallized during heating. In a basic thermal hehavior investigation
of i-poly(R,S-epichlorohydrin), Janeczek et al.20 reported the existence of three melting
peaks in the thermogram of samples crystallized at 60 oC. Although varying heating rate
studies were not performed on the samples, the results of a WAX analysis exc\uded the
possibility of polymorphism.
3.4.3.2 Comment on the Ratio of the Lower to Higher Temperature Peaks
At this point, it is appropriate to comment on the relative abundance of dominant
and subsidiary lamellae in each pcIymer sample at different crystallization temperatures,
and among the different pclymers at a given crystallization temperature. At high
3-36
Thermal Behavior
crystalli7.ation temperatures (low undercoolings), the dominant lower temperature peak in
the thermograms of all of the samples in Figure 3.7 suggests an increased amount of
branching and/or ftlling-in between the relatively few dominant lamellae in the spherulitic
structure. Accordingly, the consistently larger, lower temperature peak in the
thermograms of the i-PRSECH stereoblock polymer probably arises due to a greater
proportion of subsidiary lamellae in the spherulite, and a reflection of the relatively slower
rate of overall crystallization for this PECH form. To this end, a cross-sectionallook at
Figure 3.7 may he helpful. At a given crystallization temperature, the shape of the
i-PRSECH thermogram might he typical of an early stage of crystallization of the
polyenantiomer or the equimolar polyenantiomer blend. Likewise, the equimolar
polyenantiomer blend thermogram might depict the appearance of the polyenantiomer
thermogram after a shorter period of crystailization. The data in Figure 3.l2(b) indicate
different rates of development of the area under the lower and higher melting peaks,
however, it is not known how these relative rates change with increasing crystallization
temperature.
3.4.3.3 Remark ".>n the Conditions of the Hoffman-Weeks Plot
The melting peak temperatures in Figure 3.7, observed as a function of isothermal
crystallization temperature, were used to conslrUct the Hoffman-Weeks (H-W) plot
(Figure 2.15) in Chapter Two. The higher temperature peak in the endotherms of the
incompletely isothermally crystallized samples (Figure 3.7) may he considered to he the
melting temperature of the crystallites with a lamellar thickness characteristic of the
crystailization temperature, and not of further lamellar thickening. Thus, problems of non
linearity in the H-W plot associated with recrystailization and long-lime isothermal
thickening34 are alleviated.
It is particularly interesting to note that, had the choice been made to conslrUct the
H-W plot of data from the melting thermograms with heaÛJ'g rates of 5 instead of
3-37
•Thermal 8eilal'Îor
20 oC/min, a different, perhaps smaller linear region of the H-W plot would have Ix.'en
available for the extrapolation of the data to the predicted equilihrium melting
temperature. The reason for this is as follows: It can he said that for isothermally
crystallized PECH samples at high temperatures (ca. 110 OC), the positions of the lower
and higher temperalUre peaks in the subsequent melting thermogram ail.' œlatively heating
rate independent. Recall, however, that the thermogram of a sample which has hœn
isothermally crystallized at 90 oC and subsequently heated at a rate of 20 oC/min, exhihiL~
a double melting peak endotherm. Yet, if this sample is heated at a rate of 5 oC/min, a
third melting peak appears, arising l'rom the continuous melting and recrystallization of the
original second melting peak. Effectively, the position of the original second melting peak
is shifted to a lower temperature; the position of the new third melting peak occurs at an
even higher temperature. Consequently, the position of the highest temperature peak
would not he collinear with the high temperature peak data on the H-W plot lor the
samples crystallized at higher Tc's, where the peak position is heating rate independent.
The observation of a low temperature positive deviation Irom the linear melting
peak temperature used to extrapolate to equilibrium conditions, is a very common, almost
standard observation in many polymer thermal investigations. The lower temperature
region can he described as one in which the observed melting temperature is essentially
constant with crystallization temperature, and consequentiy results in a relatively !lat
region above the Tm =Tc line. Mandelkern et al.34 have explained this relative invariance
of Tm with Tc as fol1ows: The samples that are crystallized in this temperature region must
traverse a temperature level, on cooling, where the crystallizalion rate is extremely rapid,
resulting in the formation of crystallites of the same, small thickness. Above this
temperature region, Le. in the linear portion of the H-W plot, the complications related to
rapid crystallization are alleviated.
ln the present investigation, the phenomenon of recrystallization dominates the
melting process associated with rapid crystallization; a relative invariance of melting
3-38
•Thermal Behavior
tcmperature with heating rate is associated with crystallization at high tcmperatures.
Thus, it may bc said that at sorne intennediate isothennal crystallization temperature, the
multiple sources of multiple melting meel
3.4.4 Note on the Replacement of the Chlorine Atom in PECH
The increase in chain flexibility which results from the replacement of the chlorine
by the hydrogen atom in the conversion of PECH to PPrO is r~flected in the relatively
greater overall rates of crystallization observed for the chemically derived polymer. PPrO
is not as amenable as PECH to the study of multiple melting bchavior, since the kinetics of
crystallization and, presumably, recrystallization are significantly higher and are not as
easily resolved under nonnal experlmental conditions. Isothennally crystallized samples
for very short periods of lime (ca. 5-20 min), however, yield multiple melting
thermograms upon subsequent heating at a scanning rate of 10 oC/min.
In light of the forgoing discussion of PECH, the following can bc speculated about
the appearance of the melting thennograms of i-PRSPrO in Figure 3.9. Heating
subsequent to crystallization at the largest undercooling (Tc = 20 oC) produces a
thennogram in which the flfst peak can bc attributed to the melting of subsidiary lamellae.
There is evidence that dominant lamellae have recrystallized upon heating by the
exothennic deviation just before the large and broad third endothennic peak. The
thennogram of a sample crystallized at a smaller undercooling (Tc =30 oC) maintains the
presence of the flfst peak, and there is evidence, from a small middle peak, that the
original dominant lamellae have not completely recrystallized (during the OSC scan) into
the material represented by the final endothenn, which remains unchanged in peak
position. The thennogram of a sample crystallized at 35 oC also contains the fust peak
which always occurs at a temperature slightly above the original crystallization
temperature. Less of the crystallites represented by the middle melting peak, which is
apparently slightly shifted to a higher temperature, have sufficient lime to recrystallize
3-39
•Thermal Behal'ior
during the 10 oC/min heating scan. Fmally, at the smallest undercooling employed (Tc =
40 oC), the degree of crystallinity developed in the sample is rclatively low, and it is
presumed that the absence of the frrst peak in the thermogram is due to the dclayed onset
of the secondary crystallization of these lamellac.
3.4.5 Concluding Comments on Multiple Melting
In general, multiple peaks in the melting endotherm of a polymer can arise l'rom
different sources. They can be representative of the melting of structures of dilTering
order or crystal lattice system, present in the sample belore thermal analysis.
Altematively, multiple peaks in the melting endotherm can arise l'rom exolhermic
processes occurring during heating. The latter is a manifestation of the effecl~ of anncal
type conditions of the thermal analysis scan. In principle, the detection of such adynamie
mechanism of competing processes can he successful oruy by analysis on a similar time
frame. It is not surprising then, that PECH is most amenable for the study of the
phenomenon of multiple melting. Most likely, the phenomenon of multiple melting is
characteristic of most, if not ail semicrystalline polymers. The characteristic rates of
crystallization and reorganization of PECH are weil within the normal practical analysi~
conditions of the OSC. Perhaps it is not by coincidence that the preponderance of
multiple melting studies have been performed on polymer systems with similar (slow) rates
of crystailization, i.e., on the order of normal OSC scanning conditions, such as
i-polystyrene and PET.
3.5 Summary and Conclusions
The observed rate of the overall bulk crystallization in PECH is greatest for the
polyenantiomers, wlùch is slightly Iùgher than their equimolar blend and much greater than
3·40
Thermal Behavior
the stereoblock polymer. In general, the slow crystallization kinetics of PECH make this
polymer particularly arnenable ta the study of multiple melting behavior. The observed
multiple melting behavior in PECH has been explained in terms of two sources, depending
on the thermal history of the sarnple and the heating rate: (1) Upon melting subsequent to
crystallization at relatively large undercoolings, the original crystallites are subject to
recrystallization at slow heating rates (ca. ~ 10 OC/min), and a triple melting peak
endotherm results. The middle and highest melting peaks are strongly dependent on
heating rate. The results of partial heating experiments of the polyenantiomer demonstrate
clearly that only one crystal lattice system exists for the polymer. (2) At much smaller
undercoolings, the appearance of the thermograrn is more sensitive to crystallization lime
due to the excessively slow isothermal crystallization kinetics, and the processes of
primary and secondary crystallization can be resolved. The drive to recrystallize, during
heating, of the original high-melting crystallites is minimal, if not zero. Therefore, there is
no larnellar thickening occurring upon heating and the highest melting peak is absent; only
a double melting peak endotherm is observed.
Ullimately, the observed differences in the overall bulk crystallization kinetics
arnong ail of the PECH polymers must be related to their structural differences. Narnely,
the enantiomeric relationship between the components in the equimolar polyenantiomer
blend, and to the nature of the defect sites in the stereoblock polymer. The proposed
increase in branching of the dominant larnellar frarnework, Le., formation of subsidiary
larnellae, in the i-PRSECH stereoblock relative to the other polymers can be accounted for
by the inherent dl'fect sites in the polymer chain which covalently link the stereosequences
of opposite sense. Specifically, upon encounter of a defect site during crystallization, an
opposite-sense stereosegment of the chain rnay present a less readily crystallizable
component and consequently become rejected. This material can then crystallize later at a
slower rate as a consequence of being "trapped" between the dominant larnellae.
3-41
•Therma/ Behavior
3.6 References
(1) Wunderlich, B. Therma/ Ana/ysis; Academie: London. 1990; p.188.
(2) Runt, J.; Harrison. I. A. In Methods of Experimenta/ Physics; Fava. R. A.. Ed.;
Academie: New York. 1980; Vol. 16. Part B.
(3) Baïr, H. E.; Salovey, R.; Huseby. T. W. Po/ymer 1967. 8. 9.
(4) Ikeda. M. Chem. High Po/ym. Jpn. 1968,25(273). 87.
(5) Kanetsuna, H.; Maeda, K. J. Chem. Soc. Jpn. 1966. 69. 84.
(6) Bell. J. P.; Dumbleton, J. H. J. Po/ym. Sei., Part A-2 1969, 7. 1033.
(7) Roberts. R. C. Polymer1969, 10, 117.
(8) Yubayashi. T.; Orito, Z.; Tamada, N. Kogyo Kagaka Zasshi 1966, 69, 1798;
Chem. Abstr., 1969, 70. 5156k.
(9) Bell. J. P.; Murayama. T. J. Polym. Sei., Part A-2 1969. 7. 1059.
(10) zachmann, H. G.; Stuart. H. A. Makromo/. Chem. 1960.41, 148.
(11) Nealy, D. L.; Davis, T. G.; Kibler, C. J. J. Polym. Sei., PartA-21970, 8,2141.
(12) Roberts, R. C. J. Polym. Sei., Polym. Len. 1970, 8, 381.
(13) Holdsworth, P. J.; Turner-Jones, A. Polymer 1971,12,195.
(14) Sweet, G. E.; Bell, J. P. J. Polym. Sei., Part A-2, 1972, JO, 1273.
(15) Russell, T. P.; Koberstein, J. T. J. Polym. Sei., Polym. Phys. Ed. 1985,23, 1109.
(16) Blundell, D. J. Polymer 1987, 28, 2248.
(17) Lemstra, P. J.; Kooistra, T.; Challa, G. J. Polym. Sei., Polym. Phys. Ed. 1972, JO,
823.
(18) Lemstra, P. J.; Schouten, A. J.; Challa, G. J. Polym. Sei.: Po/ym. Phys. Ed. 1974,
12, 1565.
(19) Pelzbauer, Z.; St John Manley, R. J. Polym. Sei. A-2 1970, 8, 649.
(20) Janeczek, H.; Trzebicka, B.; Turska, E. Polym. Commun. 1987,28,123.
(21) Rim, P. B.; Runt, J. P. Macromolecules 1983, 16,762.
3-42
Thermal Behavior
• (22) Yeh, J. T.; Runt, J. J. Polym. Sei., Polym. Phys. Ed. 1989,27, 1543.
(23) Lovering, E. G.; Wooden, D. C. J. Polym. Sei., Polym. Phys. Ed. 1969, 7, 1639.
(24) Prest, W. M., Jr.; Luca, D. J. J. Appl. Phys. 1975,46,4136.
(25) Chung, J. S.; Cebe, P. Polymer 1992, 33, 2312.
(26) Chung, J. S.; Cebe, P. Polymer 1992, 33, 2325.
(27) Rodriguez-Arnold, J.; Zhang, A; Cheng, S. Z. D.; Lovinger, A.; Hsieh, E. T.;
Chu, P.; Johnson, T. W.; Honnell, K. G.; Geens, R. G.; Palacki, S. J.; Hawley, G.
R.; Welch, M. B. Polymer 1994, 35,1884.
(28) CaIdas, V.; Morin, F.; Brown, G. R. Magn. Reson. Chem. 1994,32,572.
(29) Lee, Y.; Poner, R. S. Macromolecules 1987,20, 1336.
(30) Polymer Dictionary; Alger, M. S. M.; Elsevier Applied Science: New York, 1989.
(31) Sharples, A. Introduction to Polymer Chemistry, Edward Arnold: London, 1966.
(32) AI-Raheil, J. A.; Qudah, A. M. A Polym. Int. 1995,37,249.
• (33) Nichols, M. E.; Robertson, R. E. J. Polym. Sei.; Pan B: Poiym. Phys. 1992, 30,
305. and Nichols, M. E.; Robertson, R. E. J. Polym. Sei.; Pan B: Polym. Phys.
1992,30,755.
(34) Alamo, R. G.; Viers, B. D.; Mandelkem, L. Macromolecules 1995, 28, 3205.
•3-43
CRYSTALLIZATIONKINETICS
Chapter Four
4.1 Introduction
4.1.1 Stereocomplexation Behavior in Polyenantiomer Blends
Optically active polymers are the subject of an increasing number of studies.
particularly since blends of polyenantiomers have been shown to behave differently l'rom
the component polyenantiomers.1•8 Much of this attention has been focused on the so
called stereocomplexation phenomenon. This term was fICst used by Liquori et al.9 to
describe the complementary stereospecific interactions they observed between isotactic
and syndiotaetic poly(methyl methacrylate). Separate isotactic and syndiotaetic chains of
high steric purity. formed under stereoselective polymerization conditions. co-crysta1lized
into an ordered crysta11attice with a distinctly different X-ray diffraction pattern l'rom that
of the purely isotactic sample. This along with the superior thermal stability of the former
were attributed to the formation of a stereocomplex. in which there is an interlocking of
syndiotaetic molecules in channels formed by joining the helical grooves of adjacent
isotactic helices. At that lime it was suggested that similar stereocomplexes might be
found arnong severa! sœreoregular polymers. With the increasing availability of optically
pure monomers and consequently, the synthesis of polyenantiomers. the stereocomplex
4-1
Crysta/lization Kinetics
model has becn adapted to describe a stereoselective association between optically active
stereoregular polymer chain heliœs during crystallization. On a more tangible level. one
can imagine a screw with a large thread as a cylindrical helix representing a polymer chain.
It is intuitive that two oppositely-threaded screws will fit c10ser together in a parallel
fashion than two screws with the same helical-sense thread. In the case of opposite-sense
polyenantiomer mixtures which exhibit a stereocomplex, the frequently signilicantly
improved thermal stability of the stereocomplex is derived from its racemic nature, which
allows for c10ser packing of the antichiral. or opposite sense, chains.
Ikada et a1.8have demonstrated that at various blend ratios. the polyenantiomers of
optically active poly(lactic acid) form stereocomplexes both from the melt and from
solution, that result in raœmic crystallites with side-by-side packing of polyenantiomers.
The system has becn studied in extensio including the factors that affect the formation.
morphology, phase structure, crystalline structure and degradability of the stereocomplex.
Interestingly. il was demonstrated that co-crystallization of the polyenantiomers into a
stereocomplex is restricted to blends in which the molecular weight of at least one
component polyenantiomer is helow a critical value (ca. 1 X lOS for melt and 4 x Hl4 for
solution crystallization). The crystal structure of the stereocomplex was determined by
Okihara et aI.7 to he composed of a triclinic raœmic unit œil with lateral packing of left
handed 3\ helices of one polyenantiomer and right-handed helices of the other. By
comparison, the individual optically active polyenantiomers crystallize with unique
handedness possessing a 103 helix configuration in the pseudo-orthorhombic o.-form,ll1-\2
although a less stable ~-form is composed of 3\ helices. The stereocomplex melts at a
temperature 50 oc higher than the o.-form.
Similar behavior bas becn reported by Prud'homme et a1.\.6 for the optically active
forms of poly(o.-methyl-o.-ethyl-~-propiolactone).The crystalline stereocomplex, formed
both from solution and from the mel!, possesses a Iattice structure differerlt from that of
either component polyenantiomer. The stereocomplex melts at a temperature
4-2
•
/
Crystallization Kinetics
approximately 40 oC higher than either component polyenantiomer. ln hinary hlend~ of
the polyenantiomers and in ternary blends which include the rJcemic polymer. Ihe
stereocomplex forms preferentially and controls the morphology. ln hlends rich in one
enantiomer. al: 1 complex is still formed; the excess enantiomer is claimcd 10 he Irapped
belWeen the lamellae of the spherulites of the stereocomplex and iL~ crystallization i~
hindered by the high viscosity of the partiaily crystallized medium.
4.1.2 Stereocomplexation Behavior in Products of
Stereoselective Polymerization
Because both chiral monomers are only rarely availahle lûr polymeri7.alion. Ihe
separate polyenantiomers are seldom available for blending. Consequently in the pasto
investigators of optically active polymers have relied on the nature of stereoseleclive
catalysts to yield essentially a racemic mixture of the optically active polyenantiomers hy
polymerization of the racemic monomer. A stereoselective polymerization is one in whieh
there are two active sites: one selecting and propagating the R enanliomorph and the
other, the S enantiomorph.
Dumas et al. 13 employed a stereoselective mechanism in the polymerization of
racemic monomer to prepare poly(tert-butylthiirane) composed essentially of a statistical
mixture of the R and S polyenantiomer chains. Only the levorotatory polyenantiomer was
available for comparison. The racemic product mixture was found to he more stable than
the optically active polyenantiomer, with a 40 oC difference in the melting temperature. It
was later concluded from X-ray diffraction studies14 that the optically pure polyenantiomer
crystallizes in a trigonal crystallattice with three right-handed 3. helices with statistical up
and down packing, while the racemic species co-crystallizes in a monoclinic crystal lattice.
Similarly, based on an analysis of the X-ray diffraction patterns, Sakakihara et al.u
concluded that i-poly(tert-butylethylene oxide) prepared from racemic monomer via a
stereoselective mechanism, crystallizes in a tetragonal crystal lattice with an optically
4-3
Crystallization Kinetics
inactive unit œIl. Two right-handed S helices and two left-handed R helices pass through
the unit œil existing as a random mixture of upward and downward heliœs in al: 1 ratio.
Since neither optically active polyenantiomer was available, comparison of thermal
stability or crystal structure was not made.
4.1.3 Other Levels of 'Optical Compensation'
The structural compatibility of two opposite-sense polyenantiomers in a racemic
crystal arrangement depends not only on the sense of the helix but also on the properties
of the pendent groups and their positions on the chain, as weil as on the chemical nature of
any hetero-atoms in the backbone. A stereocomplex-type crystalline structure is favored
when such an arrangement permiLS the opportunity for new forces of attraction hetween
neighboring chaIns. The striking differences in the melting temperatures reported hetween
the optically active polyesters and their equimolar blend, alluded to above, are testimony
to this effect. In the case of the polyesters, new hydrogen bonds are made between the
polyenantiomers in the stereocomplex.
The phenomenon of stereocomplexation is not universal among mixtures of
optically active polymers. Sakakihara et al.16 have classified the various W!iYS that
opposite-sense polyenantiomers may arrange in the solid state. Three such levels of
optical compensation are described here. The stereocomplexes alluded to above are ail
examples of a unit cel[ optical compensation, with a net canceling effeet of the
asymmetric chains passing through the unit œll. Altematively, if the same numher of
R- and S- polyenantiomers exisLS only at the higher structural level of the lamel\a
(crystallite) the arrangement is one which can he described by a crystallite optical
compensation. Fmally, if the two polyenantiomers do not co-crystallize into the same
lamella but, rather, form crystal1ites of ail R- or ail S- polyenantiomers, the arrangement is
termed an intercrystallite optical compensation. Sakakihara et al.16 polymerized the
raœmic propylene sulfide monomer with a stereoseleetive catalyst and optically pure
4-4
•
•
Crystalli:ation Kinetics
monomer with another coordinate-type catalyst. X-rolY diffracùon studics suggcstcd that
these polymers have the same crystal structure as evidenccd hy idcntical liber
photographs. However, the opùcally pure polyenantiomer was slightly more thcrmally
stable. Yet, il was conc1uded that the racemic polymer possesses an intercrystallite optical
compensation. Takahashi et a1.J7 have a1so reported that racemic poly(isopropylethylene
oxide) crystallizes with an intercrystallite opùcal compensation.
One of the uncertainties which can arise l'rom studies like the ahove, whcre the two
polyenantiomers are not available in their separate forros for hlending, is the distinction
between a racemic mixture of R- and S-chains, and a stereohlock polymer with long
stereospecific R- and S-segments within the same chain. It is conceivahle that the same
type of intercrystallite optical compensation can occur with a stereohlock polymer. The
optically pure lamellae can forro with opposite-sense stereospecific segmenl~ of the chain
in different lamellae. To this end, it may not be possible to distinguish hetween thcse two
crystalline forros on the basis of X-ray diffraction patterns. Yokouchi et a1. IR and
Okamura et a1. 19 have independenüy reported that the X-ray diffraction pattern of
naturally occurring optically active isotactic poly(~-hydroxybutyrate) (PHB) is
indistinguishable from that of the isotactic PHB polymerized from racemic monomer using
a stereoselective catalyst. It was concluded that the racemic polymer crystallizes into two
kinds of optically pure crysta\lites, i.e., with intercrystallite optical compensation: 16
crystallites (Iamellae) of only left-handed helices of R chains and those of only right
handed helices of S chains. The optically pure R polymer develops crystallites composed
of left-handed helices only. More recenüy, Bloembergen et a1.20•21 have reported that
physical properties of a synthetic, highly stereoregular forro of isotactic PHB
corresponded closely to those of the naturally occurring optically active polymer. A
stereoblock model was proposed with stereoregular right-handed and left-handed helical
segments of the chains contained in separate crystalline domains, Le. an intercrystallite
optical compensation. In a similar study, Yokouchi et a1. 19 a1s0 deduced an intercrystallit.:
4-5
•Crysrallizarion Kinerics
opùcal compensaùon for poly(l3-eÙlyl-~-propiolactone) prepared wiÙl a stereoselecùve
catalyst alÙlough again no comparison was made wiÙl Ùle optically acùve polyenanùomer.
4.1.4 Optically Active Poly(epichlorohydrin) & Poly(propylene oxide)
The opùcally acùve polyenantiomers of poly(epichlorohydrin) have not been
reported, and Ùlerefore Ùle crystal structures have not been determined. Several
aUÙlors22'2S have investigated Ùle crystal structure of Ùle racemic isotactic poly(R,S
epichlorohydrin) prepared wiÙl coordinate-type catalysts. as discussed in Chapter Two.
By use of polarized Iight optical microscopy. Zmudzinski et a1. 2S have shown previously
Ùlat Ùle spherulite radial growÙl rates of partially crystalline isotactic poly(R,S
epichlorohydrin) which is presumably a block copolymer. are sufficiently slow to permit
Ùle quantitative measurement over a wide range of crystallization temperatures, ranging
from just above the glass transition temperature (Tg) to Ùle melting temperature (Tm)' wiÙl
an observed maximum in Ùle vicinity of57 oC.
AlÙlough Ùle crystal structure of optically active poly(propylene oxide) has not
been reported, several descriptions have been given of Ùle X-ray diffraction pattern of
isotactic poly(R,S-propylene oxide).26-29 On Ùle basis of Ùleir studies of poly(propylene
oxide) prepared wiÙl a catalyst not known to be stereoselective. Cesari et a1.29 described
Ùle crystallattice as P2\212\' a space group requiring optically pure unit ceUs wiÙl up and
down packing of Ùle helices.28 but wiÙlout any mention of helix sense. On Ùle basis of Ùle
results of Ùlis study. oÙlers7 have subsequently included poly(propylene oxide) in the list
of polymers unable to form racemate complexes. Le.• stereocomplexes. In the on1y report
of a study of Ùle crystallization kinetics of optically active poly(propylene oxide). Magill30
found Ùlat the optically active levorotatory polyenantiomer displays overall faster
spherulite radial growth rates Ùlan Ùle racemic isotactic polymer.
4-6
•Crysrallization Kinerics
4.1.5 General Polymer Crystallization Kinetics
The rate at which polyrner spherulites grow in an isottopic medium depends
sttongly on the crystallization temperature (Tc) and can he described by the growth rate
equation of the forrn given below, developed by Hoffinan, Davis, and Lauritzen,31.32
(4.1)
•
•
where G is the spherulite radial growth rate (crn/s); Go is the pre-exponential terrn (crn/s);
l!T is the difference between the equilibriurn melting temperature (Tm0) and the isothermal
crystallization temperature (TJ, or degree of undercooling; u* is the activation energy for
transport of crystallizable segments through the melt to the site of crystallization; R is the
gas constant; T~ is a hypothetical temperature at which molecular motion associated with
viscous flow ceases; Kg is the nucleation constant; f accounts for the variation of the heat
of fusion away from the melting point and is equal t033
[2(Tc) ]
(T; +Tc )
From left to right in equation (4.1), the two exponential terrns are referred to as
the transpon and nucleation terrns, respectively, and represent the two competing
processes which govem the rate of the overall crystallization process. As the theoretical
illustration in Figure 4.1(a) demonstrates, with decreasing undercooling, the rate of
nucleation decreases whereas the rate of transport in the melt increases. The opposing
temperature dependencies of the transport and the nucleation terms lead to the maximum
in the spherulite radial growth rate at a crystallization temperature (Tmu) between the
glass transition temperature (Tg) and the equilibriurn melting temperature (Tm0)•
4-7
•Crystallizarion Kinetics
(a)
Crystallization Temperature
• Regime mj=4
Regimellj=2
Regime 1j=4
(b)
•Figure 4.1 lllustration of (a) a spherulite radial growth rate curve and (b) a theoretical
Hoffman-Lauritzen plot showing growth regimes J. II and III.
4-8
•Crysral/ization Kinelics
At relatively large undercoolings, Le. Tc < Tmax' the growth mte depends predominantly on
the variations of the transport propenies with crystallization temperature. At
undercoolings closer to the melting temperature, the growth rate is nucleation controlled.
The nucleation constant Kg in (4.1) can be further defmed as
(4.2)
•
•
where bo is the thickness of the crystallizing stem; cr and cr. are the lateral and fold surface
interfacial free energies, respectively; k is Boltzmanl1 constant; and li/Ir is the heat of fusion
per volume of monomer units, and j is the regime coefficient, dependent on the growth
regime of crystallization (defined beiow).
4.1.5.1 Growth Regimes
According to the Hoffman model,32 the nucleation-eontrolled crystallization
temperature region üf the growth rate curve can be divided into three sections, termed
growth regimes, ear;h with a characteristic mechanism of growth at the molecular level,
depending on the degree of undercooling and the rate of nucleation.34 The different
growth regimes can be envisioned as resulting from the relative rates of the two processes
at the lamellar growth face: nucleation of a new surface layer (secondary nucleation), and
subsequent lateral growth in the niches created by the secondary nuclei to complete the
layer (surface spreading). At the smallest undercoolings, Le., at crystallization
temperatures closest to the melting point, the overall radial growth is very slow, and the
rate of nucleation is small. In this temperature region, the rate of secondary nucleation is
less than that of surface spreading, indicative of growth in regime I.32,35 Growth follows a
layer-by-layer mechanism on the lamellar growth face which extends radially into the melt
After the deposition of a nuclei on the lamellar growth front, the growth step sweeps
4-9
•
•
•
Crystallizalion Kinetics
across the face of the lamella, completely covering the lamellar surface by subsequent
folding of the nucleated chains onto the substrate in adjacent positions, before a new layer
is initiated. The direction of the chain folds is perpendicular to the overall radial growth
direction.
When the undercooling is increased further, crystaBization may enter into growth
regime II, where the rate of secondary nucleation approaches the rate of surface
spreading, and multiple nucleation may occur. Nucleation of new growth steps starts a
second lamellar surface layer before the completion of the fust one by lateral growth. This
mechanisrn may a1low for sorne chain folding in the direction of the overall radial
growth.31 At the largest undercoolings, yet still above Tmax' crystaBization can proceed
into growth regime III, where the rate of secondary nucleation is greater than the rate of
surface spreading. This occurs when the separation betwecn neighboring nuclei
approaches the width of the molecular stems.32 Essentially, in Ibis growth regime, lamellar
growth is achieved only by successive additions of nuclei.
The growth rate equation (4.1) may be rewritten in logarithmic form:
u* KhlG + R(T
c_ T.) - mGo - Tc (t:.hf (4.3)
1Plotting the left-hand side of equation (4.3) as a function of should give a
Tc (t:.T)f
straight line with an intercept equal to log Go. and the product of lateral and fold surface
interfacial free energies, (Jcr., can be estimated from the slope, -Kg. If crystallization
occurs by multiple growth regimes. then the regime transition is marked by a change in the
slope. This is depicted in the theoretical illustration in Figure 4.1(b). Bach linear portion
of the curve is n-eated independently employing the appropriate value of j. As the values
of the slopes of the different regions of the curve differ only in the value of j, the relation
of the slopes among regimes 1 : II : m should he 2 : 1 : 2.
4-10
•
•
•
Crystallizeuion Kinetics
4.1.6 The Present Study
Despite the growing interest in optically active polymers, very little attention has
been directed toward the kinetics of the crystallization in blends of these chiral species.
Any stereoselection/rejection processes that occur during crystallization should manifest
their effect clearly in the kinetics. In this chapter are presented the isothennai melt
crystallization kinetics of poly(R-epichlorohydrin) (PRECH), poly(S-epichlorohydrin)
(PSECH), their equimolar blend, and the corresponding stereoblock poly(R ,S
epichlorohydrin} (PRSECH), studied using polarized light video microscopy. The effect
of the replacement of the chlorine atom on poly(epichlorohydrin} with a hydrogen atom is
investigated in a similar analysis of the crystallization kinetics of the chemically derived
poly(S-propylene oxide} (PSPrO), poly(R-propylene oxide} (PRPrO) and the stereoblock
poly(R,S-propylene oxide} (PRSPrO). The linear spherulite radial growth rates of all the
polymers studied are analyzed in terrns of the classical Hoffinan-Lauritzen treatment.
The work described in this chapter was perforrned with the aim of dcveloping a
better understanding of the crystallization mechanisms which govem the growth of
polymer spherulites from the melt through the study of the crystallization kinetics of the
above mentioned optically active polymers. It was of particular interest to detetrnine the
extent to which co-crystallization of polyenantiomers could occur and the effects of
polyenantiomer segregation on the rates of crystallization.
4.2 Experimental
4.2.1 Materials
The synthesis and characterization of the polymers used for the studies described
in this chapter have been detailed in Chapter Two.
4-11
Crystallization Kinetics
4.2.2 Polarized Light Video Microscopy
4.2.2.1 Apparatu8
The crystallization behavior of the spherulites was followed by polarized light
video microscopy using a Nikon Optiphot-Polarized light microscope in tandem with a
Linkam THMS600 hot stage, TMS91 temperature controller. and CSI96 cooling unit. A
COHU video camera mounted on the microscope transmitted real-lime images through a
computer to a secondary monitor (Electrohome ROB) for display. The computer housed
a frame-grabber (PCVisions Plus) to digilize the analog signal. These images were
simultaneously recorded onto videotape using a Mitsubishi U80 video cassette recorder
for later analysis.
Polarized light photomicrographs were taken with a 35 mm Nikon camera
mounted in place of the video camera.
4.2.2.2 Sample Preparation
The samples were prepared for viewing by ftrst pressing the polymer between two
circular glass cover slips on a hot plate at a temperature of 175 cC for
poly(epich10rohydrin) and 150 cC for po1y(propy1ene oxide) until the po1ymer just melted.
The samples were then transferred to a hot stage and me1ted for 15 min at the same
temperature under a nitrogen atmosphere, followed by cooling at a nominal rate of
130 cC/min to the se1ected isothermal crystallization temperature which Was held constant
to within ± 0.1 cC. Samp1e thickness was controlled by the use of 12 I1m-thick a1uminum
foil spacers. Only samples of 10-20 I1m thickness were emp10yed. Typically one po1ymer
samp1e was used to determine the radial growÙl rate (G) at Ùlree different crystallization
temperatures at large undercoolings, but for smaller undercoolings a samp1e was used only
once.
4-12
•
•
•
Crysral/izarion Kinetics
4.2.2.3 Spherulite Radial Growth Rate Measurements
A video analysis software program (lAVA, Jandel Scientific) was used for the
radial measurements of the digitized spherulites. For each crystallization temperature, the
radii of three spherulites in the field were measured at six regular lime intervals, weil
before impingement, during the isothermal crystallization. At each time interval, the
radius of each spherulite was taken as the average of 12-16 measurements. The
procedure was repeated for the same crystallization temperature using a fresh polymer
sample. The average slope of the resultant six straight line (r2 ~.999) plots of radius as a
function of time was taken a~ the radial growth rate (G) at that crystallization temperature.
4.2.2.4 Measurement of Spherullte Band Perlod
The combined width of a bright and dark concentric ring (band) in a banded
spherulite was taken as the spherulite band period. This distance was measured by
drawing a radial line on the digitized banded spherulite which extended over
approximately 10 band periods, or a total of 20 bright and dark rings. This distance was
divided by the actual number of band periods, which could be counted easily by
inspection.
4.3 Results
4.3.1 General Spherulite Morphology
The differences in morphology arnong the polymers in question are noted as
observed under the polarized light optical Irjcroscope with crossed polars prior to the
measurement of the radial growth rates.
4-13
•
•
•
Crysrallization Kinetics
4.3.1.1 Poly(eplchlorohydrln)
lsotacùc poly{epichlorohydrin) crystallizes from the melt readily forming
negaùvely birefringent spherulites with the characterisùc Maitese eXÙDcùon cross visible
under a polarized light microscope with crossed polars. The photomicrographs in Figure
4.2{a) through (d) illustrate this morphology, capturing spherulites of poly{R
epichlorohydrin) (PRECH), poly{S-epichiorohydrin) (PSECH), i-poly{R,S
epichlorohydrin) (i-PRSECH), and the equimolar polyenanùomer blend, respecùvely, ail
isothermally crystallized at the sarne undercooling. The spherulites of PRECH and
PSECH exhibit periodic birefringent eXÙDction rings, or bands, when the crystallizaùon
temperature exceeds the temperature of maximum growth rate (Tmax)' Interestingly, the
stereoblock i-PRSECH and the equimolar blend show no evidence of banded spherulites al
any crystallization temperature. Instead, the spherulites appear to have a coarser, more
open texture than those of the opticaily pure polyenantiomers. When the spherulites are
grown in relatively thin sections of the sarnple mel!, dendritic structures typically develop
in the equimolar blend and the i-PRSECH sarnples as shown in Figures 4.3(a) and (b),
respectively. The effect on spherulite appearance of placing a polymer crumb of the
opposite-sense PECH polyenantiomers on the microscope coyer slip and aIlowing the
mells to physically mix was investigated. The spherulites in the photomicrograph
contained in Figure 4.4 demonstrate this effect. The IWO small spherulites on the right
appear like those of the pure polyenantiomer. The spherulite on the lef!, however, is
typical of a 'mixed spherulite'. There is a banded region as we11 as a non-banded region.
The appearance of the latter is similar to the spherulites of the equimolar polyenantiomer
blend or the stereoblock shown in Figure 4.2.
4.3.1.2 Poly(propylene oxlde)
Figure 4.5 demonstrates the morpholQgy of the poly(propylene oxide) (PPrO)
polymers as observed under the polarized light opticai microscope. In similar fashion to
4-14
•
a
Figure 4.2
Crysta/lizalion Kinetics
~ ;."~,
'"" ~ \'
.~.'-:'~\'-,." \"
'.,\ ,.",
,
20 J.1rn
Polarized light optical pholomicrographs of spherulites of (a) opticallyactive PRECH, (scale bar applies to all pholomicrographs in Figure 4.2);
4-15
Crysrallization Kinetics
b
Figure 4.2 (cont'd) Polarized light optical photomicrographs of spherulites
(h) optically active PSECH,
4-16
Crysta/lizaIion Kinttics
c
Figure 4.2 (cont'd) Polarized light optical photomicrographs of spherulites of (c) the
stereoblock i-PRSECH.
4-17
Crystallization Kinetics
d
Figure 4.2 (cont'd) Polarized light optical photomicrographs of spherulites of (d) the
equimolar blend of the polyenantiomers.
4-18
Crystallization Kinttics
a 100 I!m
b SOl!m
Figure 4.3 Polarized light optical photomicrographs of dendrites of (a) the equimoIarblend of PRECH and PSECH polyenantiomers; (b) the tlereoblock iPRSECH, melt-crystallized at 80 oC in thin sample sectiOIl<l.
4-19
•Cryslallizarion Kinerics
10 J.1rn
Figure 4.4 PolariL::d light optical photomicrograph of a physical mixture of PSECH
and PRECH polyenantit)l1.icr~ melted on the same glass cover slip.
4-20
•
•
•
Crysrallizlltioll Killerics
their precursor PECH polymers. the optically pure polyenantiomers. poly(S-propylene
oxide) (PSPrO) and poly(R-propylene oxide) (PRPrO) in Figures 4.5(a) and (hl.
respectively. form banded spherulites over a wide range of temperatures ahove the
temperature of maximum growth rate (Tmu)' ln contra.~t to the corresponding PECH
polymers. the stereoblock i-PRSPrO and the equimolar hlend of the PPrO polyenantiomers
in Figures 4.5(c) and (d), respectively. crystallize from the melt forming banded spherulites
which are indistinguishable from those of the optically pure polyenantiomers. ln addition.
the spherulites of the equimolar blend of PPrO polyenantiomers do not crystallize l'rom the
melt with a uniform growth front. The occurrence of "pleats" in the spherulite gives rise
to an uneven spherulite border. This phenomenon is characteristic of the spherulites of the
equimolar polyenantiomer blend at ail the crystallization temperalUres invcstigated.
4.3.1.3 The Temperature Dependence of Band Period
The temperalUre dependence of the spherulite band period for PECH and PPrO L~
shown in Figure 4.6. The presence of banding is common in many polymer spherulites,
with the size of the extinctiun intervaI, or band period, increasing with temperature.30,37.47
The banding and generaI morphology of the spherulites of a 95:5 or 5:95 molar blend of
the polyenantiomers of PECH are indistinguishable from those of either opticaIly active
component at the same crystaIIization temperature. Periodic extinction band~ are aIso
observed in the spherulites of a 70:30 or 30:70 molar blend of the polyenantiomers but the
band period is greater than that of the spherulites of the opticaIly pure polyenantiomer at a
given temperature.
It is weIl known that the apparent texture of polymer spherulites crystaIIized from
the melt at progressively smaller undercoolings becomes graduaIly coarser. In banded
spherulites, this results in a graduai distortion of the regular, concentric birefringent
ê;xtinction Iiattem until a type of mixed, unrefined birefringence is fmaIIy apparent in the
spherulires fanned at crystaIIization remperatures close to the melting temperature.
4-21
•Crysrallizarion Kinerics
a50 J.l.rn
Figure 4.5 Polarized light optical photomierographs of isothermally melt-crystallized
spherulites of (a) optically active PSPrO (Tc = 40 OC), (scale bar applies toall photomierographs in Figure 4.5);
4-22
•Crysra/lizarion Kinnics
b
Figure 4.5 (cont'd) Polarized Iight optical photomicrographs of isothermally melt
crystallized spherulites of (b) PRPrO (Tc = 31 OC), .
4-23
•Crystallization Kinetics
c
Figure 4.5 (cont'dl Polarized light optical photomicrographs of isothermally melt
crystallized spherulites of (c) the stereoblock i-PRSPrO (Tc = 35 OC),
4-24
•Crysta/lization Kinetics
d
Figure 4.5 (conl'd) Polarized Iighl optical pholomicrographs of isothcnnally mllll
crystallized spherulites of (d) the equimolar blend of pSPrO and PRPrO
polyenantiomers (Tc = 32 oC).
4-25
•Crystallizarion Kinetics
18
16 - •14 -
,-..,
E 12 -::1.'-'"'0 10 - fil0'1::aJ •Q.. 8 -
"'0
t:-"§6 -/Xl .0
t:-4 - t:- at:- t:- t:- t:-
Éla 02 - a a
0 1 1 • 1 , • •
90 80 70 60 50 40 30
Undercooling (oC)
Figure 4.6 (a) Relationship between the spherolite band period and undercooling for
the PRECH and PSECH polyenantiomers (0), the 95:5 blend of the
polyenantiomers (0), and the 70:30 blend of the polyenantiomers (A), and
i-PRSPrO (e).
4-26
•Crysrallizarion Kinetics
The consequences of this morphology are reflecled in the hand period measuremenl~ for
both PECH and PPrO polymers in Figure 4.6. AI an undereooling of ca. 45 oC. Ihe hands
of Ihe PECH spherulites become 100 dislOrled 10 measurc. However. Ihe hand~ of Ihe
i-PPtO spherulites remain very regular unùl an undercooling of 32 oC. where Ihe hand
period is significantly large.
The influence of temperature on the spherulite hand period is hesl il\uslmled wilh
the polarize.d light photomicrograph of the PRPtO slep-growth spherulite in Figure 4.7.
The PRPrO spherulite was nuc\eated at a temperature of approximatcly 35 oC
(undercooling = 47 OC), and subsequently heated to a cryslallization lempcralurc of 53 oc.
At this crystallizaùon temperature, coarse, broad bands typically form; lhese arc visihle
near the center of the spherulite. The sarnple was then cooled 10 lhe cryslallizaùon
temperatures of 43 oC, 33 oC, and 23 oC, and isothermally cryslalli:red al each. The fasl
coolings are marked on the spherulile by abrupt changes in lhe band period.
4.3.2 Spherulite Radial Growth Rates
4.3.2.1 Poly(epichlorohydrin)
The linear spherulite radial growth rates of all of the PECH sarnples sludied are
presented in Figures 4.8 and 4.9 as a function of erystallizaùon temperalure. There is no
difference between the growth rates ofPRECH and PSECH polyenantiomers. The growth
rates of a 95:5 polyenantiomer blend are the sarne as those of the individual
polyenantiomers, however the growth rates of a 70:30 polyenantiomer blend are
decreased. The growth rates of the equimolar (50:50) polyer' -.tiomer blend are depressed
over a wide range of temperatures relative to those of the component polyenantiomers.
The entire cryslallization curve for the stereoblock i-PRSECH is maIkedly reduced relative
to that of PRECH or PSECH. For ail of the polymers the maximum in the growth rate
occurs at the sarne erystallization temperature, Tmu' equal to 63 oC, corresponding to a
4-27
Cryslalliza1ion Kinetics
100 Ilm
Figure 4.7 Polarized light optieal photomierograph of a PRPrO step growth spheruliteshowing the intluenee of temperature on the spherulite band period. Thespecifie erystallization temperatures an: noted in the texl
4-28
•Crysrallbllion Kinerics
18 _r_-----------------,
O+-~..........___r_,.._r_"'"T"""_r_..___.,__,.__.___r___.__r_...---...__l
20 30 40 50 60 70 80 90 100 110
Crystallization Temperature (OC)
2
4
16
•
Figure 4.8 Plot of the spherulite radial growth rates dependence on isotherrnal
crystallization temperature for the 95:5 (V), the 70:30 (O). and the 50:50
(0) molar blend of PSECH and PRECH polyenantiomers.
4-29
Crysta/lization Kinetics
18 -r-----------------,
16
2
o -+-...--r-........~,...._I""""'1,......,__r........,.___r'_....__r"""'T""_r=.....,r_1
20 30 40 50 60 70 80 90 100 110
Crystallization Temperature (OC)
Figure 4.9 Plot of the spherolite linear radial growth rates dependence on isothennalcrystal1ization temperature for PSECH and PRECH polyenantiomers (e),
the equimolar polyenantiomer blend (0), and the stereoblock i-PRSECH
(.).
4-30
•
•
•
Crysta/lization Kinetics
Tnw.rrm0 ratio of 0.82, which is in excellent agreement with that found for other crystalline
polymers.36,37
It is of particular interest that the growth rates of the all of the PECH polymers can
00 measured over a wide temperature range, and therefore the complete data analysis is
warranted. The experimentally determined growth rates of PRECH, PSECH, the
equimolar polyenantiomer blend, and the i-PRSECH stereoblock were analyzed in terms of
the Hoffman-Lauritzen treatrnent. The raw data was fit to the growth rate equation (4.1)
using a Gaussian function of a peak-fitting software program (PEAKFIT, landel
Scientific). The terms Kg and U* were permitted to vary, maximizing the correlation
coefficient. In the analysis of the PECH growth rate data, the experimentally determined
values of Tg and Tm0 were used, -26 oC and 138 oC, respectively. The value of T_ was
chosen by definition3l as Tg-30 oC. In the present analyses the regime coefficient, j, was
chosen as 4 (the assignment is evaluated in the discussion). The thickness of the growth
step, bo' was taken as the spacing OOtween the lIO planes of symmetty, as in
polyethylene,3l and is equal to 4.54 x 10-8cm for PECH.22 From the unit œil density22 of
1.47g/cm3, and enthalpy of fusion of a perfectly crystalline sample,38 143.6 l/g, Âilr was
determined to 00 2.11 x 109 erg/cm3•
The oost-fit estimate of U* was used to conslrUct the Hoffman-Lauritzen plot in
Figure 4.10, from which the oost-fit estimates of the parameters Go and Kg were
determined. It can 00 seen that the data for each of the polyenantiomers, their equimolar
blend, and the stereoblock of PECH fall on straight lines in Figure 4.10, indicating
crystallization within one regime. AlI of the OOst·fit estimates including aac are listed in
Table 4.1.
4-31
•Crystal/ization Kinetics
4.03.02.0
-7.0 -1-.......-...--..-........--................,...-,-................-...-..--...---l
1.0
-2.0 -.-----------------,
,.-...E--.
-3.0E--.'"'-"
~M0
-4.0MN
~+~ -5.0<1'-"
1
\jb{) -6.0:3
( 2.303 Tc (ôT)!) x 105
Figure 4.10 Hoffman-Lauritzen plots for PRECH and PSECH (e), the equimolar
polyenantiomer blend (0). and the stereoblock i-PRSECH (.).
4-32
•
•
•
Crysral/ization Kinetics
Table 4.1 Best-fit Estimates of the Growth Rate Paramelers for PECH
Go U' Kg aa'cpolymer (cm/s) (k1/mol) (x lOS K2) (erg2/cm4)
i-PRSECH 0.12 ± 0.07 7.3 ±0.3 1.55 ± 0.06 603
PRECH & PSECH 0.07±0.03 7.0±0.3 1.28± 0.04 494
eguimolar R & S blend 0.04±0.02 6.7+0.2 1.26±0.04 494
a Calculated using j = 4.
4.3.2.2 Poly(propylene oxlde)
The characteristic uneven border of the spherulites of the equimolar
polyenantiomer blend make reproducible radial measurements difficult to perform. Figure
4.11(a) demonstrates the irregular growth of the spherulitic front at an undercooling of
51°C (Tc = 31 oC). Two significantly different growth rates are measured. The slower
and faster growth rates are determined measuring the center and the outside of the pleat.
respectively, as indicated in Figure 4.11(b). The two radial growth rates measured at this
crystal1ization temperature fall in between the values of corresponding radial growth rates
for the polyenantiomer and the stereoblock polymers. The equimolar blend of the PROO
and PSPrO polyenantiomers does not form part of the complete crystallization kinetics
analysis.
To clearly show the effect of dechlorination of PECH on the resulting growth rate
kinetics. the spherulite radial growth rates of the polyenantiomers and the stereoblock
forms of each of PECH and POO as a function of undercooling are presented in Figure
4.12. The overall growth rates of the dechlorinated polyether (POO) are increased by an
order of magnitude. As in the PECH system, the growth rates of i-PRSPrO are depressed
relative to those of PSPrO or PRPrO over the spectrum of crystal1ization temperatures
srudied.
4-33
•Clysrallizarion Kinerics
2.01.61.20.80.4
20 +-........-.-....---...........""T"'""...........,....-,--.-
0.0
70b
60..-..e:::i.'-'
<I:l 50::s.-"0~
~
~.- 40-2eu
• .c::0.
en30
Time (min)
•Figure 4.11 (a) Optical photomicrograph of a typical spherulite of the equimolar blend
of PRPrO and PSPrO polyenantiomers showing pleat formation and
(b) The ~ftect of the non-uniform border of the spherulite in (a) on the
measurement of the spherulite radial growth rate. (Tc =31°C)
4-34
•Crystal/harhm Kinnics
14 ,.••
"'"' 12 • •~ o~oJl \t)
10'-"
~8 \
~ \~ 8i 00 •
o •
~ 6 0
e·o·
• ~ cg •• po( •-g 4 ~.
~ ~•2 0 .......-..
• aŒJ 0 a:JI:PIIl• a0
120 100 80 60 40 20
Undercooling (OC)
•
Figure 4.12 Spherulite radial growth rate dependence on undercooling of PROO.
PSPrO (e). and i-PRSPrO (0). PRECH. PSECH (->. and i-PRSECH (0)
showing the effect of dechlorination of PECH to POO on Cl)'stallization
kinetics.
4-35
•
•
•
Crysrallizalion Kinetics
The experimentally determined linear radial growth rates of PSPrO, PRPrO and
i-PRSPrO (shown in Figure 4.13) were also fit to equation (4.1) using the following
values. The experimentally determined values of Tg and Tm0 were used, -65 oC and 82 oC,
respectively. The value of T_ was taken by definition as Tg-30 °C.31 The regime
coefficient was chosen as 4 (the choice of j is discussed later). The width of a growth
stem, bo, was taken as the spacing belWeen the 110 planes of syrnrnetry,39 equal to
4.24 x 10-8 cm.28 From the unit celI density,28 1.097 g/cm3, and enthalpy of fusion data,zg
804 kJ/mole, tihr was determined to be 1.58 x 109erg/cm3.
The parameters U*, Go' and Kg were allowed to vary during the fit to equation
(4.1) in order to maximize the correlation coefficient The best-fit estimate of U* for
i-PRSPrO (5400 kJ/rnol) was used to conslIUct the Hoffman-Lauritzen plots of ail of the
polymers in Figure 4.13, from which the values of Go and K8
were estimated from the
intercept and slope, respectively, and are listed in Table 4.2. However, as Figure 4.13
demonstrates, il is not possible to fit all of the data of either the polyenantiomers or
i-PRSPrO to the predicted straight line relationship. Inste3.d, each system displays a
change of slope in the same high temperature region (Tc ~ 37 OC).
Table 4.2 8est·fit Estimates of the Growth Rate Parameters for PPrO
portion of Go U* K. Cl"Q·,
polymer the curve (x 1O-2cm/s) (k1/rnl'1) (x lOS K2) (erg2/cm4)
;-PRSPrO upper 1.2±OA 5.4 ± 0.2" 1.42±0.07 517
lower 0.07±0.03 1.04±0.05 380
PRPrO& upper 0.8 ± 0.4 6.2 ± 0.6 1.31 ±0.07 478
PSPrO lower 0.09±0.04 1.03 ±0.05 375
"The estimate of 5.4 kJ/rnol was used to estimate Go and Kg (and cm.) for bompolymers.
4-36
•Crysta/lization Kint'tics
-1(:1) (b)
-2 • upper • upper0 lower 0 lower
,......., -3,......,
E--.8
1 -4E--.
u
'-"~
-5........
?;+ -6Cjt::~ -7
-800
-9
4 6 8 104 6 8 lalI[Tc(L\1)fl x 105
Figure 4.13 Hoffman-Lauritzen plots for (a) PRPrO and PSPrO polyenantiomers and
(b) the stereoblock j-PRSPrO.
4-37
•
•
•
Crystallization Kinetics
4.4 Discussion
4.4.1 Spherulite Radial Growth Rates: Poly(epichlorohydrin)
At large undercoolings, where ttansport processes are rate-controlling, the
difference OOlWeen the growth rates of PRECH or PSECH and those of the equirnolar
polyenanùomer blend gradually diminishes. This is expected since the energy required to
transport chains through the melt of the opùcally pure polymer would 00 the same as that
of the equimolar blend. The oost-fit esùmate of the u* parameter is essenùally the same
for PRECH, PSECH, the equimolar polyenanùomer blend, and i-PRSECH, at a value of
7.0 x 103 J/mol. Since u* reflects the mobility of the chains, wlûch is reflected by the
value of Tg, its invariance among these polymer systems with idenùcal glass transiùon
temperatures is expected. This calculated esùmate is similar to the empirical univers::!
value of 6280 J/mole suggested by Hoffman.40
The pre-exponenùal parameter (G.) may 00 the most poorly defined parameter of
the growth rate equaùon, mainly due to a generaIlack of growth rate data for polymers at
large undercoolings. The observed differences in the derived values of Go reponed in
Table 4.1 are relaùvely small, as might 00 expected for polymers that are .0 similar in
structure.
The derived value of the nucleation constant (Kg) is 1.27 X lOS 1(2 for PRECH,
PSECH, and the equimolar polyenantiomer blend, but il is significantly larger for the
i-PRSECH stereoblock polymer, at 1.55 x lOS 1(2. The difference can 00 attributed to a
different contribution of the lateraI and fold surface free energy product (0"0".) or to a
different regime coefficient (J).
Molecular variables, such as the presence of defect sites on a stereoblock polymer
chain, presumably can affect the rates of secondary nucleation and surface spreading and
thus influence the type of growth regime that is operational during crystallization. The
straight line dependence in Figure 4.10 for each of the PECH polyenantiomers, the
4-38
•Crysrallharion Kirlerics
equimolar polyenantiomer blend, and the stereoblock givcs no indication of a rcgim.:
transition over the crystallization temperalUre range. However, the possihility that th.:
ahove disparity in the Kg estimalCs is due to the crystallization of th.: stercohlock polym.:r
in a constant, but differenr growth regime l'rom th.: other polymers must he consid.:rcd.
4.4.1.1 Assignment of the Regime Coefficient (J)
Table 4.3 Iists the different values of crcr, which result l'rom ail of the possiol.:
assignments of j for PRECH, PSECH, the equimolar polyenantiomer blend, and
i-PRSECH.
Table 4,3 Estimates of crcr. for Ali Possible Values ofj for PECH
V"CfOOblock' crcr, (erg2/cm4)
case jpoly,...tiomer,blend) i-PRSECH PRECH, PSECH & Blend
(i) (4,2) 603 988
(ü) (2,4) 1206 494
(ili) (2,2) 1206 988
(iv) (4,4) 603 494
The results of crcr, for each case in Table 4.3 are evaluated. ln principle, the value
ofj accounts for the change in slope of a multiple regi...e Hoffman-Lauritzcn (II-L) curve
and therefore maintains the value of crcr, relatively constant throughout ail of the growth
regimes. However, in case (i) and (ü) above, the choice of a different value of j for
i-PRSECH relative to the other polymers has the effeet of creating a large difference in the
respective crcr. values. ln addition, if i-PRSECH is crystallizing in a different growth
regime over the entire temperature range relative to the other polymers, it follows from
the relative slopes of the H-L curves that j =4 for i-PRSECH and 2 for the other
4-39
•Crystallization Kinetics
polymers [from Figure 4.J(b)] with similar slopes. Thus case (ü) is invalidated. Based on
a value of ca. 1180 erg2/cm4 for polyethylene,41 it can also he said that a value of 1206
[case (ii) and (iii)] for PECH in general is an overestimate. Finally, in case (iv), the
assignmcnt of a common value ofj =4 for ail of the polymers is deemed rea~onahle.
4.4.1.2 Lateral and Fold Surface Free Energy (O"O"J
Case (iv) above does not only arise by the eIimination of the other cases hut, the
variance hetween the estimates of 603 and 494 erg2/cm4 for i-PRSECH and the other
polymers. respectively, can specifically account for the disparity in the respective Kg terms
according to the following: It is proposed that the defect sites, covalently linking the
stercoregular sequences of the i-PRSECH chain, are interfering with the process of regular
chain folding. Assuming that molecules attach themselves to the growing lamellar surface
one segment at a lime,42 the growth will continue along the substrate until a defect site is
cncountered, at which lime regular chain folding coulà he in!errupted, or halted. In this
case, loose chain loops and chain ends would develop on the surface and increase the free
energy of the fold surface of the i-PRSECH lamella. It is known that for a given polymer,
the value of 0"0", increases with decreasing isotacticity, indicating an increase of roughness
for bath lateral and fold crystal surfaces.43 The lateral or fold surface free energy, or both,
may constitute the difference in the 0"0", values for the polymers. Ta estimate the specific
contributions of the each surface Cree energy the following calculation was performed.
According ta the relation,31
(4.4)
where lX is an empirical parameter that is assigned a value of 0.1 for polyethylene and
other polyolefins,44 the estimate of the lateral surface free energies for the PECH polymers
4-40
•CrystalIi:ation Kinerics
is a constant 9.6 erglcm2• This in tum gives an estimate of 63 erglcm2 for the fold surface
free energy of i-PRSECH and 51 erglcm2 for each of PRECH. PSECH, and the equimolar
polyenantiomer blend. The a and a. estimates are compared to those of other polymers
and POO in the next section.
It can he said that the larger value of aa. for the stereohlock rel1ecL~ the increased
size and number of loops atthe fold surface that are caused hy the defect sites. Moreover.
since the native defect in the chain introduces an opposite-sense stereoregular segment.
lateral growth may he retarded as the substrate accommodates the r.ew chain segment.
This accommodation may be realized through branching or splitting which are mcchanisms
of sphemlite development introduced by certain types of defecL~ on crystal growth
surfaces which can stop the lateral crystal growth. Excessive branching or splitting of thc
lamellae would have a di;ninishing effect on the overall mdial growth of the crystal. This
mechanism can account for the observed depression of the stereoblock growth rates over
the entire crystallization temperature range as compared to those of either polyenantiomer
or the equimolar polyenantiomer blend.
4.4.2 Spherulite Radial Growth Rates: Poly(propylene oxide)
In contrast to PECH. the overall radial growth rates of the i-PRSPrO stereoblock
polymer are depressed by only approximately 30 % relative to fue polyenantiomers over
the crystallization temperature range. Table 4.4 contains very similar best-fit estimate.~ of
the growth rate parameters for i-PRSPrO and the polyenantiomers. It is evidcnt from the
appearance of the (H-L) curves in Figure 4.13 that the various PPrO polymers are
crystallizing in a common growth regime. However. the question which arises is whether
the break in the slope of each polymer constitutes a regime transition. This question is
addressed systematically by evaluating each case of possible j values for PPrO in Table 4.4
helow.
4-41
Crystallization Kinetics
• Table 4.4 Estimates of cc. for Ali Possible Values ofj for PPrO
portion of (jstereoblock' cc. (erg2/cm4)
case the curve jpolyenantiomer) i-PRSPrO PRPrO. PSPrO
(i) upper (2.2) 1034 956
lower (4,4) 380 375
(ü) upper (4,4) 517 478
lower (2.2) 760 750
(iü) upper (2,2) 1034 956
lower (2,2) 760 750
(iv) upper (4,4) 517 478
lower (4,4) 380 375
4.4.2.1 Assignment of the Regime Coefficient (j)
The terms upper and 10wer refer, respectively, to the regions above and helow the
change in the slope of the H-L curve. Case (i) can be easily judged incorrect. Values of 2
and 4 for j for the upper and lower regions. respectively. of the H-L curve denote a
regime 1111 transition, where in principle. the slope of the regime II (upper) portion (Of a
curve is less than that of the regime 1 (lower) porti':>n [from Figure 4.1(b)]. The opposite
is observed for all of the polymers in Figure 4.13. In case (ü) the values ofj are chosen to
rel1ect '4 regime II/III transition. which at a fust glance, more accurately represents the
relative slopes of the upper and lower regions of each curve in Figure 4.13. However. the
ratios of the upper to lower Kg terms for i-PRSPrO and the polyenantiomers are 1: 1.33
and 1:1.27. respectively. According to theory. the j values can only he effective in
maintaining a constant cc. value throughout bath regimes when this ratio is close or equal
to 1:2. In case (ü). the choice of j amplifies the düferences in cc. and consequenüy. is
deemed incorrect. It must he noted, however. that there are reports of multiple regime
hehavior in structurally similar polyethers. e.g. poly(oxymethylene),32.4s and poly(ethylene
oxide).46
4-42
•Crysralli~arion Kineric.\'
Altemaùvely. the break in the slopes of the H-L cUives can he considcred a
deviation of the high temperature data l'rom a single regime crystallizing polymcr system.
Before discussing cases (üi) and (iv). the implicaùons of a deviation l'rom single regime
growth are considered. The change in slope of the H-L curves occurring at Tc> 37 oC in
the PPrO polymers could he related ta the distribuùon of the molecuiar wcights in the
crystallizing polymer. Wunderlich has extensively reviewed the pher.omenon of molecular
segregaùon when crystallizing a polydisperse system at smail undercoolings.4N Others4')
have reported such preferenùal crystallizaùon of isotacùc polypropylene chains of high
molecular weight at high Tc.
Recall that Kg also depends on the thickness of the adding lamellar surface layer
(bo)' If the growth face of a crystallizing lamella changes in a particular temperature
region, such a shift would be manifested as a discontinuity in the slope of the H-L curve in
that temperature region. Unless the distance between the growth planes for the two active
lamellar growth faces are related by a factor of two, the relaùon of the Kg values above
and helow the slope discontinuity will not he characteristic of a classical growth regime
transiùon. This mechanism of multiple growth faces has heen invoked by sorne to accou:lt
for the proposed regime WIll transition in poly(ethylene oxide).47 It i~ interesùng to note,
however, that according to the crystallographic data reported for i-PRSPrO in Chapter
Two, the interplanar distance for the (most intense) 110 rel1ecùon. taken as bl) in defining
Kg, is 1.22 times the size of the interplanar distance corresponding to the second mosl
intense rel1ection.
In the single growth regime case (ili), the esùmates of 1034 and 956 erg2/cm4 for
i-PRSPrO anu the polyenantiomers, respectively, which result from a choice of j =2, arc
relatively large in comparison with the values helow, reported for other polymers. In case
(iv). however, the estimates of 517 and 478 erg2/cm4 although very similar. relate weil
with those values estimated for PECH reportetl herein and for other structurally similar
polymers. The lateral and fold surface free energy producl has been estimated in the range
4-43
•Crysta/lization Kinetics
of 496 erg2lcm4 to 627 erg2/cm4 for poly(ethylene oxide),lO a value of 733 erg2/cm4 ha~
hcen estimatcd for poly(lactic acid),ll and an estimate of 635 erg2/cm4 ha~ hcen rcported
for poly(E-caprolactone).l2
The values of the individuallateral and fold surface free energies for i-PRSPrO and
the polyenantiomer.~ can he estimatcd using the crcr. (upper) estimates l'rom case (iv) in
equation (4.4), and employing the empirical value of 0.1 for lX. A common value of 6.7
erg/cm2is deterrnined for cr. This value is quite low compared to that found for the PECH
polymers (!1.6 erg/cm2) and is a reflection of the relatively lower !!J.hl for PPrO. The cr. is
therefore estimated to he 77 and 71 erg/cm2 for i-PRSPrO and the polyenantiomers,
respcctively. The relatively higher estimates of cr. for PPrO compared to those for the
corresponding PECH polymers may indicate that the replacement of the pendent carbon
chlorine atom on PECH with the hydrogen atom generates more l1exibility in the chain
which can consequently forrn tight folds. Fold surface free energy estimates for other
polymers include 40-50 erg/cm2 for poly(ethylene oxide),SO 61 erg/cm2 for poly(lactic
acid),SI and 46 erg/cm2for poly(~-hydroxybutyrate).S3
Magill30 has also estimatcd similar crcr. values for both the optically active
po'.yenantiomer (M. = 10 300) and the R,S-polymer (M. = 135 (00) of PPrO. The results,
rnnging from 190 to 250 erg2/cm4 for both polymers, were also deterrnined from an
analysis of the spherulite radial growth rates. However the radial growth rates are
relatively slow compared to those found in the present investigation (ca. 30 % reduction
for each polymer). It is difficult to compare these estimates of acre with the ones reported
in the present investigation. In the analysis by Magill, Tm° was taken as the observed
optic . 'Iting point (ca. 75 OC) and an underestimated /!,hlo value (ca. 14 x 108 erg/cm3)
was used based on that of the structurally similar polymer, poly(ethylene oxide). The
relatively slower growth rates of the PRSPrO polymer in the Magill study may he a
reflection of regioirregularities present in the chain. Defects of this type would
presumably pose a greater hindrance, relative to stereo defects, to the polymer during
4-44
•Crystalli~ation Killerics
crystallization. Indeed Magill rcported that the nonhanded spherulites pllssessed faster
growth rates than the rarely ohserved handed spherulites in t"e same PRSPrO polymer.
The specilic angle of rotation was not reported for the PRSPrO polymer so thal it is no!
known if it was truly racemic. It would seem. however. lhat Ihere was at kasl a sm.ùl
fraction of the polymer which may have hecn "enantiomerically rich" which would accounl
for the observation of the banded spherulites.
The results of the present study suggest that the lamellae of the slereohlock and
the polyenantiomers of PPrO have similar surface free energies. Thus. in keeping with the
above statements made for PECH. the defect sites in the i-PRSPrO stereohlock also
introduce an opposite-sense stereoregular segment. and normal growth may he retarded.
however to a lesser extent. as the substrate accommodates the new chain segment. The
mechanism of accommodation in the i-PRSPrO stereohlock polymer may involve less
branching.
4.4.3 Morphology and the Crystallization Kinetics
The presence of banding in polymer spherulites is generally ascribed to a twisling
of crystallographic orientation al -'lt radii that apparently rel1ecl:. a cooperative twisting of
radiating larneUar crystals about their axes of fastest growth. and implic.ï a high degree of
coordination in the packing of the larneUae.54-58 Polarized light microscopy and light
scattering studies of banded spherulites have assigned the observed extinction bands and
diffraction rings to the helicoidal orientation of the crystals within the spherulites.s~.6o
Similar work has correlated the periodicity of the extinction band~ with X-ray
microdiffraction and electron diffraction patterns of the banded and non-bandcd areas of
the spherulite. presenting direct evidence for larnellar twisting.55,56 The suhject of the
possible origins of banding is the focus of Chapter Five and will not he discussed further
here.
4-45
•Crysrallization Kinetics
The banding and general spherulite morphology of a 95:5 blend of the PECH
polyenantiomers are indisùnguishable from those of either opùcally acÙve component at
the same crystallizaùon temperature. If this 5% addition is considered an "impurity" in the
otherwise pure homopolymer system, the insensiùvity witb respect to spherulite
morpbology is in contrast with the dramaùc effect of the lesser 3 % defect site "impurity"
in the i-PRSECH stcreoblock polymer (esùmated in Chapter Two) which displays coarse,
open spberulites.61 This is consistent with a stcreoblock structure and emphasizes the
important role that these backbone defects play in the crystallite growth. As the increase
in band period with inereasing blend ratio demonstrates [Figure 4.7(a)], the em~ct of
blending of the polyenantiomers on the spherulite morphology is to make the spherulites
more open,56 increasing the size of the bands. Blending of PRECH and PSECH apparently
hinders the twisùng of lamellae as would occur naturally in the component polyenanùomer
at a given crystallizaùon temperature, or else relieves the stress on the lamellae which
causes them to twiSt.40,60,62 Fmally, at a polyenanùomer blend ratio of 50:50, the effeet of
one PECH polyenanùomer on the other is to apparently abolish regular larnellar twisùng,
as evidenced by the nonbanded spherulitcs at ail crystallizaùon temperatures ~~ :eIl as the
nonbanded dendriùc structures visible at small undercoolings. Regarding the absence of
banding in the i-PRSECH spheruIitcs, it would seem that the effect of the inherent defect
sites between stereosequences is to disrupt the regularity of the chain to a sufficient extent
that any regular twisting of the lamella is inhibited at ail crystallizaùon temperatures.
The mixed morphology spherulitc in Figure 4.5 is remarkably similar to an example
of a mixcd type IIII spherulitc reported by Dreyfuss63 in a study of a series of crystalline
poly(R.S-epichlorohydrin) polymers with varying degrees of optical activity. Dreyfuss
describcd two types of spheruIitcs, type 1 and type II, the latter possessing a fmer fibrillar
structure and only in rare cases displaying a banded morphology. Occasionally, both
forras could be seen in the same spherulitc. Type II spherulitcs were observed to melt at a
slightly higher temperature, crystallized more rapidly, and had sorne optical activity. Il
4-46
•
•
•
Crysrallizarioll Killeûcs
was concluded from that qualitative study that type Il spheru\ites were fonned from
optically active polymer sequences while type 1 spheru\ites were fonned fonn rncemic
polymers with \iule or no optical activity.
In keeping with the above statements for PECH, it seems that the effect of the
stereo defect sites on the regular lamellar twisting of i-PRSPrO is neg\igible, based on the
similar morphologies of i-PRSPrO and the polyenantiomers. Magi1l3o reponed that a
banded spherulite appearance was rare in the case of the PRSPrO. However, it must be
noted that the stereoblock polymer in the present is not truly racemic, but has an optical
purity of ca. 37 % (Chapter Two).
4.5 Summary &Conclusions
The individual polyenantiomers of PECH display indistinguishable spherulite
morphologies as weil as spherulite growth rate kinetics. The PRECH and PSECH
polyenantiomers develop a banded spherulite morphology while the stereoblock
i-PRSECH polymer and the polyenantiomer equimolar blend have nonbanded, coarser,
open spherulites. The radial growth rates of the equimolar blend of the polyenantiomers
are overall depressed (by -60 %) relative to those of either componenl The radial growth
rates of the stereoblock i-PRSECH are markedly reduced compared to all of the polymers.
A oost-fit analysis of the observed growth rate data to the Hoffman-Lauritzen growth rate
equation indicates a common rate of transport across the phase boundary for ail of the
PECH polymers. The nucleation pararneter, Kg, however, is greatest for the stereoblock
i-PRSECH polymer. The growth rate is very sensitive to changes in this exponent; the
rates of Ctystallization of the i-PRSECH polymer are consequently dramatically less than
those of the optically active PRECH and PSECH polyenantiomers or the equimolar
polyenantiomer blend, indicating the significant role of the defect site during crystal
growth.
4-47
•
•
•
Crystallization Kinetics
The small reduction in the radial growth rates of the equimolar polyenantiomer
blend relative to either the component PRECH or PSECH polyenantiomer can he
attributed to the rejection of opposite sense chains at the growth front of stereospecific
lamellae. A stereospecific segregation is expected to induce branching of the lamellae with
the rejection of opposite sense chains, thus inhibiting the regular twisting of the lamellae
and causing a coarse, open spherulite to develop. A similar mechanism can occur during
the crystallization of the stereoblock i-PRSECH polymer, and would involve the rejection
of opposite sense stereoregular segments. Since the stereoregular segments are covalently
linked, the lamellae would he excessively branched and limited in their size. Again, the
coarser, nonbanded, open spherulites of the stereoblock polymer support this type of
stereospecific segregation at the growth front. The proposed formation of stereospecific
lamellae in melt crystallized spherulites of the equimolar polyenantiomer blend and the
stereoblock PECH polymers has heen described in other systems as an intercrystallite
optical compensation, and precludes the existence of stereocomplexation.
The effect of substitution of the chlorine atom in PECH with a hydrogen atom is
reflected in an order of magnitude increase in the growth rates over the entire range of
crystallization temperatures in the resulting PPrO polymers. The radial growth rates of
each of the polyenantiomers of PPrO are tlle same. The radial growth rates of the
stereoblock i-PRSPrO are depressed (by -30 %) relative to that of the PRPrO and PSPrO
polyenantiomers over the complete range of crystallization temperatures studied. The
polyenantiomers, the equimolar polyenantiomer blend, and the stereoblock form of PPrO
al! develop banded spherulites at crystallization temperatures above the temperature of
maximum growtIJ rate (Tmax)' The results suggest tlJat tlJe dechlorination of PECH, to
form PPrO, greatly influences tlJe mechanism of accommodation of tlJe opposite-sense
stereosequence in tlJe i-PPrO stereoblock. The relatively small reduction observed in tlJe
growth rates hetween tlJe stereoblock and tlJe polyenantiomer stereosequence in tlJe
stereoblock influences tlJe sense of tlJe stereoblock helix and controls tlJe overall
4-48
•Cryswlli:arioIJ KiIJ,'rics
morphology. This differenee may ret1eel inereased t1exihilily. as shown hy IhL' dem:ase in
Tg.
4.6 References
(1) Grenier, D.; Prud'homme, R. E. Lcborgne, A; Spassky, N. J. PoIYIII. Sc'i. 1981,
19,1781.
(2) Grenier, D.; Prud'homme, R. E. J. Polym Sei. 1984,22,577.
(3) Voyer, R.; Prud'homme, R. E. PolYIII. Prepr.(AIII. Chelll. Soc.. Dil'. PoIYIII.
Chem.) 1988, 29, 611.
(4) Riteey, A M.; Prud'homme, R. E. Macromolecules 1992, 25, 972.
(5) Riteey, A M.; Brisson, J.; Prud'homme, R. E. Macromolecules 1992, 25, 2705.
(6) Riteey, A M.; Prud'homme, R. E. Macromolecules 1993,26, 1376.
(7) Oldhara, T.; Tsuji, M.; Kawaguchi, A.; Katayama, K.; Tsuji, H.; Hyon, S.-H.;
Ikada, Y. J. Macromol. Sei.• Phys. 1991, B30, 119.
(8) Tsuji, H.; Ikada, Y. Macromolecules 1S93, 26,6918. and previous articles in lhis
series.
(9) Liquori, A. M.; Anzuino, G.; Corio, V. W.; D'Alagni, M.; De Sanlis, P.; Savino.
M. Nalllre 1965, 206, 358.
(10) Marega, C.; Marigo, A.; Di Noto, V.; Zannetti, R. Makromol. Chem 1992. 193,
1599.
(11) Kovacs, A. J.; De Santis, P. Biopolymers 1968. 6, 299.
(12) Hoogsteen, W.; Postema, A. R.; Pennings, A. J.; ten Brinke, G.; Zugenmaier, P.
Macromolecules 1990, 23, 634.
(13) Dumas, P.; Spassky, N.; Sigwalt, P. Makromol. Chem 1972, 156. 55.
4-49
Crystalli~arion Kim'tics
(14) Matsubayashi. H.; Chatani, Y.; Tadokoro, H.; Dumas. P.; Spassky, N.; Sigwalt. P.
Macromolecules 1977, JO, 996.
(15) Sakaldhara, H.; Takahashi, Y.; Tadokoro, H.; Oguni, N.; Tani. H.
Macromolecules 1973, 6. 205.
(16) Sakakihara. H.; Takahashi. Y.; Tadokoro, H.; Sigwalt, P.; Spassky, N.
Macromolecules 1969, 2, 515.
(17) Takahashi, Y.; Tadokoro. H.; Hirano, T.; Sato, A.; Tsruta, T. J. Polym. Sei.,
Polym. Phys. Ed. 1975. 13, 285.
(18) Yokouchi. M.; Chatani. Y.; Tadokoro, H.; Teranishi, K.; Tani. H. Polymer 1973.
14.267.
(19) Okamura. K.; Marchessault, R. H. In Conformation of Biopolymers;
Ramachandran. G. N., Ed.; Academie; London. 1967; vol. 2. p. 709.
(20) Bloembergen, S.; Holden, D. A; Bluhm, T. L.; Hamer, G. K.; Marchessault, R. H.
Macromolecules 1989 22,1656.
(21) Bloembergen, S.; Holden, D. A; Bluhm, T. L.; Hamer, G. K.; Marchessault, R. H.
Macromolecules 198922, 1663.
(22) Richards, J. R. Ph.D. Thesis. University of Pennsylvania. Philadelphia. 1961. Diss.
Abstr. 1961 22. 1029.
(23) Perego. G.; Cesari, M. Makromol. Chem. 1970, 133, 133.
(24) Hughes, R. E.; Cella, R. J. Polym. Prepr. (Am. Chem. Soc.• Div. Polym. Chem.)
1974, 15, 137.
(25) Zmudzinski, L.; Sokol. M.; Turska, E. Acta Polym. 1985,36,483.
(26) Natta, D. G.; Corradini, P.; DalI'Asta, G, Arri. Accad. Naz. Lincei. Cl. Sei. Fis.
Mat. Nat. Rend. 1956, 20, 408.
(27) Shambelan, C.; Hughes. R. E. Natl. Meet.-Am. Chem. Soc. 1958.
(28) Stanley, E.; Litt. M. J. Polym. Sei. 1960,43.453.
(29) Cesari, M.; Perego, G.; Marconi, W. Makromol. Chem. 1966, 94,194.
4-50
•Crysrallization Kinerics
(0) Magill. J. H. MakronwL. Chem 1965, 86. 283.
(31) Hoffman. J. D.; Davis. G. T.; Lauritzen. 1. 1.. Jr. In Treatise on SaUd Stare
Chemistry; Hannay. N. B.. Ed.; Plenum Press: New York. 1976; Vol. 3. Chap. 7.
(32) Hoffman, 1. D. Polymer 1983. 24. 3.
(33) Ong. C. J.; Priee. F. P. J. Polym. Sei. Polym. Symp.. 1979. 63. 59.
(34) Lovinger. A.1.; Davis. D. D.; Padden. F. J.• lr. Polymer 1985. 26.1595.
(35) Hoffman.l. D. Polymer 1985. 26. 803.
(36) Mande1kem, L. Crystailization of Polymers; McGraw-HiIl Co.: New York. 1964;
Chapter 8, p. 264.
(37) Godovski, Yu. K. Polym Sei. USSR. 1969, Il, 2423.
(38) laneczek. H.; Tr.rebicka. B.; Turska. E. Polym. Commun. 1987,28. 123.
(39) Chanzy, H.; persona1 communication, December, 1995.
(40) Hoffman.l. D.; Lauritzen, 1. 1.. Ir. J. Res. Nat. Bur. Standards 1961. A65. 297.
(41) Clark, E.l.; Hoffman, J. D. Macronwlecules 1984, 17. 878.
(42) Keith, H. D.; Padden, F,1., Ir. J. Appl. Polym Phys. 1964, 35. 1286.
(43) lanimak, 1.1.; Cheng, S. Z. D.; Giusti, P. A. Macronwlecules 1991, 24, 2253.
(44) Lauritzen,1. 1.. Jr.; Hoffman, J. D. J. Appl. Phys. 1973.44,4340.
(45) Pe1zbauer, z.; Galeski, A. J. Polym Sei. 1972, C38, 23.
(46) Cheng, S. Z. D.; Chen. J.; Janimak.1. J. Polymer 1990, 31, 1018.
(47) Marentette, J. M. Ph.D. Thesis. McGill University. Montreal, Canada. 1995. See
also Point, 1. 1.; Damman, P.; Janimak, 1.1. Polymer 1993, 34. 3771.
(48) Wunderlich, B. Faraday Disc. R. Soc. Chem 1979, 68, 239.
(49) Wang. Y. F.; Lloyd, D. R. Polymer 1993, 34, 2324.
(50) Alfonso. G. C.; Russell, T. P. Macronwlecules 1986, 19, 1143.
(51) Vasanthakwnari, R.; Pennings, A, 1. Polymer 1983, 24, 15.
(52) ~hillips. P. J.; Rensch. G. J.; Taylor, K. D. J. Polym Sei.: Pan 8: Polym Phys.
1987,25,1725.
4-51
Crysralli:ari(lll Killerics
(53) Pierce. R.; Brown. G. R.; Marchessault. R. H. P(llymt'l" 1994. 35. WH4.
(54) Priee. F. P. J. Polym Sei. 1959.39. 139.
(55) Keller, A. J. Polym Sei. 1959,39, 151.
(56) Fujiwara. Y. J. App/. Polym. Sei. 1960, 9, 10.
(57) Bassel!. D. c.; Hodge, A. M. Polymer, 1978, 19.469.
(58) Keith. H. D.; Padden, F. 1., Jr.; Russell, T. P. Macromolecules, 1989.22,666.
(59) Li, W.; Yan, R.; Jiang, B. Po/ymer 1992, 33, 889.
(60) Morra, B. S.; Stein, R. S. Po/ym Eng. Sei. 1984, 24, 311.
(61) Keith. H. D.; Padden. F. 1., Jr. J. App/. Phys. 1963,34.2409.
(62) Keith, H. D.; Padden, F. J., Jr. Po/ymer 1984, 25, 28.
(63) Dreyfuss, P. ACS Symp. Ser. 1975,6,70.
4-52
•SPHERULITE
MORPHOLOGY
Chapter Five
5.1 Introduction
Il is weU known thal folded-chain lameUae are the building blocks of melt
crystallized polymers. However, the detai\ed description of the geometry of individual
lamellae and their organization in the multilayered crystal aggregates !bat forro spherulites
remains the subject of intense study. Traditionally, effons to describe the fundamental
features of lameUar organization in melt-erystallized polymers have been focused on
polymers with a minimum degree of complexity, principally, polyethylene.1-12
Consequently, the models that have been developed rely heavily on these rather ideal
systems.
Detai\ed investigations have also been made of the morphology of isotactic
polypropylene, i-PP, which is a more complex crystalline polymer owing to the pseudo
chiral nature of the chainS. I3-19 The a-monoclinic crystal forro of i-PP is among the most
thoroughly studied and hence best understood polymer crystal\ine forros. Chiral selection
of the pseudo-ehiral helices leads to the most stable phase of this crystal\ine forro, in which
!here exist regularly altemating layers of right- and left-handed helices.
5-\
•Spheruiir,· Morphoiogy
The extent to which the effecl~ of true asymmetric ccmers along the polymer
backbone can be transmitted to higher-order structures has received limited attention)" and
invites further investigation. Macromolecular backhone chirality can impose characleristic
secondary structures and. in turn. influence the mode of packing upon cryslallization.
Specifically, it can he envisioned that the effect of hackhone chirality on thc organization
of moleclllar helices in the lamellae and. hence. the arrangement of neighboring lamellae.
would he martifested in the overall morphology of lhe cryslalline structure (e.g. a
spherulite). In this way. the examination of the morphology of the spherulites of these
optically active polymers could provide new insighl~ into the organization of lhe lamellae
within spherulites. A search of the Iiterature to dale, however. failed to provide any report
of a study which employs optical activity as an analytical tool for the elucidation of the
spherulitic architecture of melt-crystallized synthetic polymers.
Poly(epichlorohydrin) is a polymer with truc chiral centers along il~ polyether
backbone. In Chapter Four it was shown that on r.rystallization from the melt, the
optically pure polymer forms spherulites which, when examined by polarized light optical
microscopy, exhibit a periodic pattern of birefringent extinction bands while the
corresponding equimolar blend of the polyenantiomers forms nonbanded spherulites.
5.1.1 Banding in Polymer Spherulites
The appearance of periodic birefringent extinction bands under transmitted
polarized light is a relatively common feature in spherulites of crystalline polymers. It is
weil established that il is a regularly twisted molecular orientation about the radial growth
axis that gives rise to this pattern.3-4.22.24 However, a twisted molecular arrangement can
he accommodated in more than one type of lamellar structure. The precise description of
the shape and organization of the lamellae in these banded spherulites remains a subject of
ongoing debate. Polyethylene,7'9 a substituted polyethylene, poly(4-methyl pentene-I),25
and the more complex poly(vinylidene fluoride)26 are among the few systems in which the
5-2
•Spherulite Morph%gy
geometry and the mutual disposiùons of lamellae in banded spherulites have been
examined in dctail.
5.1.1.1 Current Models
Thcte currcntly exist two models which descrihe the fonn and organizaùon of the
lamellac within banded spherulites: (1) A twisted molecular geometry can he expressed at
the lamellar level through a sequence of transverse screw dislocaùons of the same sign.
Undcr isothermal crystallization condiùons these are equally spaced along the radial
growth direction of the lamella. as depicted in Figure 5.1 below.
•Radial Growth
•Figure 5.1 A schemaùc representation of a lamella possessing a series of isochiral
screw dislocations along its length (taken from reference 27).
5·3
•Spht'rlllile Morph%gy
Each screw dislocation supplies an increment of the over.ùl twist ln the mdially
growing lamellae. The sense of the isochiml transverse dislocations is controlled hy the
direction of inclination of the chain stems with respect to the lamellar normal and the
nature of the folding at the surface of an individual lamella. According to lhis model. the
overall effect of neighboring. in-phase, dislocating lamellac gives rise III the ohserved
banding pattern. This mode!, originally put forth hy Schultz and Kinloch.27 has since hœn
modified and developed extensively by Bassett and coworkers.6•N•2N Currently the model
holds that the chiraliry, or sense, of the dislocations is linked to the direction of tilt of the
chain stems and is not associated with any stresses at the lamellar rold surraces. At the
screw dislocations, which serve as branch points in the spherulite, the adjacent lamellae
which constitute the layers of the screw dislocation mutually diverge, enhancing the
element of twist. This diversion is attributed to the pressure, belWeen layers, due to
uncrystallized molecular cilla associated with the fold surfaces or neighboring lamellac,
i.e., there is an interlamellar origin of the lamellar geometry.
(2) A regularly twisted molecular orientation can also he accommodated in a
continuously twisted helicoidal lamella without screw dislocations, as originally proposed
by Keller.29 As the results of this study are discussed mainly within this model, the
illustration of a helicoidal lamella is contained in the discussion section. The twisting i~
attributed mainly to the influence of the surface stresses arising l'rom the disordered chain
folds. 3o This model has been developed significantly by Keith and Padden.N The rold
staggering causes the chain stems to he non-orthogonal to the fold surfaces. The non
orthogonal relationship between the lamellar faces and the chain axes creates unequal
stresses at opposite fold surfaces. This, in turn, gives rise to hending moments and
ultimately results in the twisting of the lamella. Screw dislocations may result in given
regions as a consequence of contacts between growing, but already twisted, lamellae. The
feature which detennines the direction of twist, the so-called "chiral factor", is the
5-4
•Spherulite Morphoiogy
direction of chain tilt with respect to the lamellar normal. Keith and Padden have thus
focused on an intramolecular origin of the lamellar geometry.
5.1.2 Methods of Investigation
Direct observation of melt-crystallized lamellae has routinely been achieved by
electron microscopy. The application of transmission electron microscopy techniques is
limited to very thin sections or films and, to a certain extent, tends ta suffer from the
inherent effects of sample irradiation. Developments in chemical etching,31 decorating.32
and staining33 as sample preparation techniques have helped to minimize these problems.
but provide indirect observations of the structures. Indeed. sorne of the fmest images of
lamellar morphology have becn obtained through electron microscopy of surface replicas.
However, atomic force microscopy (AFM) is a non-destructive scanning microscopie
technique which allows direct observation of the original surface yielding local three
dimensional informati('n in real space. Direct imaging of bulk or film surfaces. or
microtomed sections thereof, either in air or in liquid, has becn used to study the
morphology of both crystallinel4,34 and liquid crystalline polyrners.35 In most cases, AFM
alleviates the need to extract the subject from its original matrix thus allowing in situ
polyrner studies.36 High resolution AFM has become a valuable tool in the resolution of
molecular structure37 and, in the case of i·PP. the direct observation of helix sense in a
crystal has becn reported.13
5.1.3 The Present Work
In this chapter are reported the results of a detailed investigation of the spherulite
morphology, using polarized light optical microscopy (PLOM), rel1ectance light optical
microscopy (RLOM), and atomic force microscopy (AFM), of melt-crystallized
spherulites of optically pure poly(R-epichlorohydrin), (PRECH), and poly(S
epichlorohydrin), (PSECH), and the equimolar blend of the enantiomers. It is the aim of
5-5
Spherlliilt' Morphoiogy
the present work to deterrnine the effecl~ of hackhone maeromolecular chimlity on the
lamellar assemhly within melt-crystallized. handed spherulites through an investigation of
the spherulite morphology.
5.2 Experimental
5.2.1 Materials
The polymers used in the study of spherulite morphology include poly(R
epichlorohydrin) (PRECH), poly(S-epichlorohydrin) (PSECH). their equimolar hlend. for
which the details of their synthesis and characterization have already heen given in Chapter
Two.
5.2.2 Polarized Light Optical Microscopy (PLOM)
Polarized light optical photomicrographs were ohtained using a Nikon 35 mm
camera mounted on top of a Nikon Optiphot-Polarized Light Microscope. The samples
were prepaned for viewing by PLOM using the procedure outlined in Section 4.2.2.2. The
radialline intensity prome of the polarized light optical micrograph in Figure 5.S(c) was
obtained using JAVA vide:! analysis software (Jandel Scientific).
5.2.3 Reflectance Light Optical Microscopy (RLOM)
Reflectance light optical micrographs were taken with an Olympus 35 mm camera
mounted on top of a Leitz reflectance microscope. Samples were prepared for viewing in
the same manner as outlined below for observation using AFM.
5-6
•Spherulire Morph%gy
5.2.4 Atomic Force Microscopy (J'FM)
5.2.4.1 Basic Operating Principles
The operating principle of the AFM relies on the attractive and repulsive Van der
Waals forces which exist specifically between the atoms of the probe tip and those of the
sampIe. Figure 5.2 illustrates the two operating regirnes of the AFM. In the concacr
regime, repulsive forces exist between the atoms of the probing tip and those of the
sample. In the non-contacr regime, the AFM operates with the tip-to-sample distance is
optimized to allow for the attractive Van der Waals forces to exist between the tip atoms
and the surface atoms. A1though other forces are involved in the interaction between
sample and tip (i.e. magnetic. electric, water capillary forces. and physical frictional
forces). under normal operation it is the Van der Waals forces which account principally
for the deflection of the cantilever as it probes the surface morphology of the sample.
Figure 5.3 is a schematic representation of the instrument in the contact operating
regime arrangement. The sample is mounted onto a stage. which is a piezoelectric tube.
The size of the piezoelectric tube determines the maximum extension of the scanner in the
x-y directions and therefore determines the maximum square area of the sample which can
be probed during a single scan. The stage is capable of extremely fine raster motion in the
x-y scanner plane. An x-y scan generator controls the voltage applied to the piezoelectric
stage to maintain the selected scanning frequency of the mounted sample under the
stationary probe. The probe is composed of a tip attached to the end of a flexible
cantilever with a particular square pyramidal or conical geometry. The spring constant of
the cantilever is specifically lower than the spring constant holding the sample atoms
together 50 that upon scanning the tip over the sample, the repulsive Van der Waals forces
will effect bending of the flexible cantilever. If the cantilever is too stiff and the force too
great. the sample will deform before the cantilever reacts. The change in the tip-to-sample
5-7
•
•
•Figure 5.2
:
SpIJeru/ire MorplJ%gy
non-contact AFM image
contact AFM image
Schematic illustration of the two operating regimes of the AFM.
(a) contact and (b) non-contact (taken from reference 38).
5-8
•Spherulire Morphology
deflection sensor
0 feedhack• electronics
antilever freferencesignal
V tactive display
~ample
~ .~raster motion
scanner x - y scan •generator
'- ./z plotted ingrey scale
x,y
z
c
Figure 5.3 Schematic illustration of the AFM instrument in the contact operating
regime (taken from reference 38).
5-9
•
•
•
Sphl'r/l/ir,· Morph%gy
distance as the cantilever reacL~ to the sample topography. changes the force het\Veen the
tip and sampIe atoms. as illustrated in Figure 5.3.
A laser beam directed onto the back of the cantilever is detlecled onlO a position
sensor-pholOdetector. As the eantilewr is displaced in the ~-direction upon scanning. the
position of the detlected laser on the detector is corrcspondingly displaced. generating a
slowly varying dc current. The amplified dc current is compared to a refen:nc.: and th.:
resulling error signal is directed into a z-feedback controller. The f.:.:dback loop functions
to maintain a constant deflection of the cantilever by adjusting the ~-position of th.:
sample. This is done by sending an appropriate z-scanner voltage. based on the Icedback
information. to the scanner causing a contraction or extension of the piezod.:ctric tube.
Since a constant force is maintained by the adjustment of the z-position of the sample.
scanning in this mode of contact AFM is called topo or constant force mod.:. The
feedback signal. wlùch corresponds ta the surface topography. is used ta map out the
image of the sample.
The feedback control loop can also be disabled. i.e.• no correction in the tip-to
sample distance. and the error signal used directly ta map the topographical image of the
sample. With the feedback turned off. the sample is not being raised or lowered but
remaining at a constant height under the probe. Scanning in this mode of contact AFM L~
termed error or constant height mode.
5.2.4.2 AFM Instrument and Settings Used in the Present Investigation
The AFM images were recorded in air using an AutoProbe'" CP scanning probe
microscope (Park Scientific Instruments. Sunnyvale, CA.) operating in either the constant
force or constant height mode, as indicated. using a 100 /lm scanner operating at a
scanning frequency of 2 Hz. The cantilevers employed were Microlevers'" and Sharp
Microlevers'" (park Scientific) wlùch had a nominal radius of 500 and 200 Â. respectively.
and a nominal spring constant of 0.05 N/m. Where indicated, AFM images scanned with a
5-10
•Spherulire Morph%gy
constant force were subjccted LO a high-pass mter. The AFM was used exclusivdy in the
contacl operating regime, and images were collected while scanning in both the 10pO and
lhe error modes. The samples were scanned al a raie of 2 Hz, unless otherwise indicated
on the ligure caplion of the image.
5.2.4.3 Sample Preparation
The samples were prepared for AFM by rlfSI pressing a polymer crumb hetween
Iwo silicon warer fragments (Semiconductor Processing Inc., Boslon, MA.) about 1 cm2,
on a hotplale at 175 oC until the polymer melted. The silicon warer assembly was
transferred to a Linkam THMS600 (U.K.) microscope hotstage in tandem with a Linkam
TMS91 thermal controller. The polymer was melted at 175 oC for 15 min under a
nitrogen atrnosphere, and then cooled at a nominal rate of 130 oC/min to a crystallization
lcmperaturc of 70 oC where it was held isothermally for 30 min. The silicon warer
assembly was then removed from the hotstage and immersed in liquid nitrogen. This
treatment permitted the top silicon warer to he removed easily, leaving a polymer rùm of
the desired thickness, ca. 30-50 I1m. The remaining botlom warer and polymer film were
then re-inserted into the hotstage and melted again for 15 min, cooled at a nominal rate of
130 oC/min to an isothermal crystallization temperature and allowed to crystallize with an
unrestrained top surface for 4 h under a nitrogen atmosphere.
5.2.4.4 Image Processing
The high-pass mter function included in the PSI Data Analysis Program was used
to mter the AFM images where indicated. Filtering of the real image is a convertient
method of measuring spatial periodicities in the sample. A high-pass mter was applied to
the real topo images of the spheru1itic surfaces to improve the resolution of the small
spatial periodicities observed between the spherulite bands. To apply a high-pass mter,
the raw image was fust Fourier transfonned (F. T.) inLO a two-dimensional power
5-11
•
•
•
Spherulire Morph%gy
spectrum. In this F. T. the rdative intensities of the frequency components in the y-scan
direction are on the y-axis. and the relative Întensities of the frequency components in the
x-scan direction are on the x-axis. The low-frequency contrihutions to the surface
topography. Le.. those of the larger spatial pcriodicities in the samplc. arc located near the
origin of the x- and y-axes of the power spectrum. The high-frequency contrihutions arc
extended from the origin. Filtering specitic frequency componenl~ pcrmil~ the distinction
of the large scale surface features from the fine scale roughness. In the present study. the
aim of the fùtering process was to accentuate the fme-scale lamellar edge features. hy
subtracting the large-scale banding features of the spherulite. A high-pass hand mter was
thus applied ta the real topo image allowing only those high-frequency contributions to the
surface roughness ta pass through. Through trial and error, an optimum !iller handwidth
of 30 % of the power spectrum was chosen, and the optimum slope of the cut-off region
was similarly selected as 10 %.
5.3 Results and Discussion
5.3.1 On the Three Dimensional Nature of the Spherulltes
Upon isothermal crystaIlization from the melt over a wide range of temperatures,
the optically pure polyenantiomers of poly(epichlorohydrin) form spherulites that exhibit a
banded morphology, when viewed optically by lransmitted polarized lighl As Figure
5.4(a) demonstrates, the banding pattern of the isochiral spherulites is also evident upon
microscopic viewing of the free surface by reflected light This morphology is in sharp
contrast to the nonbanded spherulites which form from the melt of the equimolar blend of
the enantiomers, under identical crystalIization conditions. The reflectance light
micrograph in Figure 5.4(b) shows the relatively coarse surface texture of the unbanded
spherulites.
5-12
•Spherulire Morphology
a
b
SOp.m
Figure 5.4 Reflectance optical micrographs of the unrestrained melt-crystallizedspherulitic fl1m crystallized at 80 oC of (a) the optically pure
polyenantiomer. PSECH. and (b) the equimolar blend.
5-13
•
•
•
Spherulire Morph%gy
The unrestrained melt-crystallized polymer lilms of the optically acùv.: enantiomers
typically contain spherulites in which the nucleus is situated in a central pit. This feature is
demonstrated clo~arly in the thn:e dimensional AFM image of a quadrant of the opùcally
pure poly(S-epichlorohydrin) spherulite sUlface shown in Figure 5.5(a). This image was
generated while scanning the surface of the spherulite in the constant force mode. with the
dark regions indicaùng depth. These pit-like depressions are of the order of 25 Ilm in
diameter and are typically of the order of 1 Ilm deep. This feature is not uncommon to
spherulites grown with a free top surfacelO.22.26.39. and presumably ret1ecl~ the exhaustion
of crystallizable material which prevents symmetry in three dimensions. Ccrtainly these
depressions are much more open than the central holes reported by Lustiger et a1. IO for
polyethylene spherulites grown in much thicker mms.
The periodic topographical pattern of ridges and valleys, clearly visible in Figure
5.5(a), is characteristic of the spherulite surface of the optically pure enantiomers of
poly(epichlorohydrin). In a previous study of banded spherulites of poly(trlmcthylenc
glutarate), Keller22 noted that in vertical illumination the bands of the spherulites were
visible as t teps or ledges on the surface of the unrestrained spherulites. The AFM image
in Figure 5.5(a), which is similar in appearance to the spherulite fragment of
poly(trimethylene glutarate) shown in Figure 7 of reference 22, demonstrates physically
the presence of such ledges a10ng the incline of the pit-like depression in the banded
spherulite of PSECH. The corresponding polarized light optical micrograph of poly(S
epichlorohydrin) is shown in Figure 5.5(h) with its sectional analysis showing the band
spacing in Figure 5.5(c). The periodicity of the birefringent extinction banding pattern
corresponds directly to the periodicity of the surface contour banding of the spherulite as
measured by AFM.
Figures 5.6(a) and (h) contain the raw, low magnification AFM images of the free
spherulitic surfaces of the S- and R-poly,~nantiomers, respectively, obtained while scanning
in the constant height mode. The coherence of the surface bands is striking. The bright
5·14
•Spherulire Morph%gy
IfiU
0,30'
0.15··
0.00
o o
5
20
15
10 /111'
Figure 5.5 (a) Three dimensional AFM image of a quadrant of an optically pure
PSECH spherulite surface crystallized unrestrained at 80 oC, scanned in the
constant force mode,
5-15
•Spherulire M(lrpllOl(l.~Y
c
,,'...,'
•oo
•
•.0..."'. •• •••
....• •. "o. 0
0..
,.".o
•00
'. 0-.•
'..,•...
• 8
6....-.-CIl= 4~-=-2
oo 20 25 30 35
Radius (microns)
•Figure 5.5 (contin'd) (h) the corresponding polarized Iight oplical micrograph. and (c)
the radialline intensity profile. showing the band periodicity, of the banded
spherulite section shown in (h).
5-16
1
a
b
Spherulite Morph%gy
o 20 40 50"",
Figure 5.6 Low magnification AFM image scanned with a constant height of (a)PSECH spheru\ite and (b) PRECH spherulite, both melt-crystallizedunrestrained at 75 oC.
S-17
•
•
Splrerulire MorpilO/ogy
bands in the AFM image indicate elevated rcgions. measurcd to lie approximately 400 Â
above the dark band regions at the highest point. The elevated bands are rcgions in which
the lamellae are oriented edge-on. These bands correspond to the bircfringent (bright)
bands under the polarized light microscope. The alternaùng surface crests and troughs of
the unrestrained banded spherulites as imaged by AFM conlirm an alternaùng llat-tIH:dge
on lamellar orientaùon, inferred l'rom similar electron micrographs of surface rcplica.~ of
similarly banded spheruiitic surfaces of polyethylem) and poly(vinylidene lluoride).26
5.3.2 The Appearance of a Surface Pattern
Figures 5.7(a) and (b) contain the corresponding raw AFM images of the arca
surrounding the nucleus of the S- and R-polyenanùomer spherulites. respecùvely. collected
while scanning the free surfaces in a constant force mode. The liltered images of Figures
5.7(a) and (b) are shown in Figures 5.7(c) and (d), respectively. Vpon close inspection of
these images, it is apparent that the lamellae, in their edge-on orientaùon. maintain a
unique sense of inclination throughout the spheruiite. A comparison of Figures 5.7(c) and
(d) leads to the observation that these lines are inclined in opposite directions for the two
polyenantiomers. The once-fI1tered images in Figures 5.7(c) and (d) were liltered once
more to remove the concentric bands l'rom the image. The resulting images are contained
in Figures 5.7(e) and (0, respectively. The image processing permits the clear observaùon
of the fmer features, Le.• the traces of the lamellar edges. and a mulù-armed spiral pattern
emerges. The prominent feature of the AFM images of the isochiral spherulitcs is the
overall appearance of a spiral, originating at the nucleus and conùnuing with a unique
sense of direction throughout the surface of the spheruiite. although it becomes somewhat
less coherent towards the edges of the larger spherulites. More striking L~ the observaùon
that the sense, or handedness of the spiral is uniquely related to the chirality of the
constituent polyenantiomer. When viewed looking down onto the free surface. the AFM
images of the banded PSECH spherulite surfaces consistently display a counter-elockwise
5-18
a
Spherulite Morph%gy
b
c
o 10 20 30 4OI'ltlO
d
10 20 30
Figure 5.7 Higher magnification AFM images of the corresponding spherulites in
Figure 5.6 scanned with a constant force of (a) PSECH, (b) PRECH, (c)
the high-pass mtered image of (a), and (d) the high-pass mtered image of
(b). Note the unique direction of inclination of the lameUar edges in each of
the isochiral spherulite surface images.
5-19
e
Spherulite Morph%gy
f
o 10 20 30 401(111 o 10 20 30 40 1'111
Figure 5.7 (contin'd) (e) The high-pass fùtered image of Figure 5.7 (c) and (f) is the
high-pass filtered image of (d). The second filter process removes the largebanding structural features and leaves the !ine lamellar edge structures.
5-20
•Spherulire Morph%gy
spiral sense. while those of the PRECH spherulite surfaces exclusively manifest the
opposite. clockwise spiral sense. The sense. or handedness. of a spherulite formed l'rom
the optically pure polyenantiomer with a characteristic nucleus pit can he determined
readily upon inspection of the surface by AFM. OnIy in rare cases was the spiral not
readily discernible. With a common viewing perspective. the handedness of the spiral i~
always opposite for spherulites of enantiomers of opposite chirality.
FiguwoS 5.8(a) and (b) contain the reflectance Iight optical micrographs of the
crude surfaces of unrestrained melt-crystallized spherulites of PSECH and PRECH
samples, respectively. The lamellar edges are not visible in these low magnification optical
micrographs. What is outstanding however. is that the bands appear not to he at constant
radius. and consequently the band Iines can he traced to form a flat spiral. The opposite
sense polyenantiomers crystallize to form spherulites which display opposite sense spirals.
In 5.8(c). however, it can be seen that the spacing between the bands becomes larger when
the two polyenantiomers begin to "mix" at the outer edges of the spherulite on the left side
of the photomicrograph. Fmally, in the middle of the same spherulite, the appearance of
the banding pattern is lost
In contrast to the present case, previous studies with non-optically active polymers
have reported that the IWO directions of spiral twist are observed to occur with equal
probability in the banded spherulites which display surface spiral morphologies.22.39• The
sense of the spiral was a1ways maintained within one spherulite or at least within an
undisturbed sector. Apan from early work by Keller,22 who fust noted that the bands of
the poly(tetramethylene glutarate) spherulites were effectively spirals, these observations
in polymer systems have principally been made for polyethylene.23-24.39b.4lla,b In those
previous investigations of the Cree surfaces, by reflected light and electron microscopy of
the surface replicas, the sense of the spiral was accented due to the inclination of the
apparently edge-on oriented lamellae in the direction of the spiral. The same inclination
phenomenon is manifested in the isochiral spherulites in this study.
5-21
•
•
•Figure 5.8
Sphtrulitt Morph%gy
a20~m
Reflectance light optical micrographs of the crude surface of meltcrystallized (a) PSECH spherulites (Tc =75 OC). Nole the !',rface spiral;
5-22
•
•
•
Sphtrolitt Morph%gy
b20J1m
Figure 5.8 (com'd) Reflectance light opticai micrographs of the crude surface of meltcrystal1ized (b) PRECH spherulites (Tc = 75 OC). Note the surface spiral;
5·23
•
•
Sphaulitf Morph%gy
•
c
Figure 5.8
20J1m
(cont'd) Reflectance light optical micrographs of the crude surface of a
physicai mixture of PRECH and PSECH (c) melted and crystallized on the
same silicon wafer (Tc =75 oC). Note the "mixed spherulite".
5·24
Spherulite Morph%gy
In those previous studies, it was suggested that the sense of a regularly twisted
lamella is developed early in the formation of the spherulite and maintained throughout the
growth.39< Mathematical models developed to account for surface spiral morphology in
polyethylene were based on the assumption that the twisted lamellae interdependently
fotm part of a continuous, predictable structure throughout the banded spherulite.41 The
mathemaùcal models arranged the twisùng lamellae in a radial fashion such that the locus
of all points of a given lamellar twist within the spherulite generated a conùnuous surface
of a thrce dimensional spiral form. In more recent work by Lustiger et al.,tO the
experimentally obseIVed surface morphology of banded polyethylene spherulites was also
accurately represented by a computer-modeled surface proj&tion of an assembly of
radially oriented same-sense twisted lamellae. Graphical representations of oblique
sections which dissect radially growing helicoids predicted different lamellar surface
projections of the model spherulite. Depending on the angle of the dissection plane to the
lamellae, and the position of the plane on the helicoidal axis, the resulting lamellar profiles
projected onto that plane varied from straight-line to S- and C-shaped lamellar geometries,
all of which can he scen experimentally for polyethylene. A sketch of one of the
computer-generated sections showing the C-shaped lamellar edges is contained in Figure
5.9. Only the C-shaped lamellar promes permitted the ready determination of the
direction of the coordinate twist of the lamellae, or the handedness of the helicoid. It was
both predicted and found experimentally that C-shaped lamellae would project onto the
spherulitic surface when the nucleus was about 5-15 helicoidal periods below. In our
CUITent study of the highly organized optically active poly(epichlorohydrin) system, the
determination of the sense of the isochiraI spherulite spiral was most easily ascertained
when the nucleus was obseIVed to he al the botlom of a pit-like depression with a radius
of about live band-periods.
The phenomenal effect on the spherulitic morphology that resulL~ from the mixing
of the two polyenantiomers to fonn an equimolar blend is demonstrated in the AFM image
5-25
•
•
Sphtrulitt Morphology
Figure 5.9 A schematic recoDSbUction of the computer-generated spherulite surface ofpolyethylene. showing the C-shaped lamellar profiles. The direction of theCs denotes the handedness of the lamellae. and the spherulite (taken fromreference 10)•
5-26
eSpherulire Morph%gy
shown in Figure 5.10. Clearly. the level of cooperativity that is required for long-range
coherence of the lamellae. in both the radial growth direction and circurnferentially. which
exists in the isochiral spherulites. is absent in the spherulites formed from the equimolar
blend. The black marks on the AFM image of the blend spherulite indicate deep holes and
cracks between sorne neighboring groups of lamellae.
The AFM images of a higher magnification afford a more detailed look at the
lamellar organization in spherulites of the PSECH polyenantiomer. Figure 5.ll(a). and of
the equimolar blend, Figure 5. Il (b). The radial growth direction is from left to right in
each image with the nucleus out of view. In (a) it is difficult to follow a particular lamella,
or group of lamellae. through a series of flat and edge-on orientations because the growth
direction of the radiating lamellae does not remain para1Iel to the viewing surface over as
many bands.24 In (b) short-range order is visible in the equimolar blend sample as many
sma11 bunches of flat, somewhat coherent lamellae. The size of a typical bunch is of the
order of 111m2.
5.3.3 Dendritic Growth
Due to sample flow during the melting period, the film thickness is reduced
considerably at the edges. Consequently, in these regions the supply of crystallizable
material is exhausted early and results in immaturely formed spherulites and dendritic
structures. Although these structures are crystallized under the same thermal conditions
as the spherulites in the melt-rich sections of the film, the former crysta1lize at a slower
rate due to the lack of crysta11izable material. The images of these regions of the optically
pure PSECH and the equimolar blend are shown in the reflected light micrographs in
Figures 5.l2(a) and (b), respectively. In the PSECH sample, these dendritic structures had
a counter-clockwise pinwheel-like appearance, reflecting the counter-clockwise twisting
sense of the mature spherulite counterpart shown earlier. These curved dendritic
structures closely resemble those of polyethylene grown under similar conditions îor which
5-27
•SpheTII/ire Morph%gy
o 10 20 30 40 SO l'IT!
Figure 5.10 AFM image of the unrestrained melt-crystallized surface of the equimolar
blend spherulite crystallized at 80 oC scanned with a constant height.
5-28
•Spherulite Morph%gy
a
n 2 31
SInn
Figure 5.11 High magnification AFM images of the unrestrained surtàce of the melt
crystallized sphemlite at 80 oC scanned with a constant height of (a)
PSECH;
5-29
•Spherulire Morph%gy
b
o 2 3 fi Inll
Figure 5.11 (cont'd) High magnification AFM images of the unrestrained surface of themelt-crystallized spherolite at 80 oC scanned with a constant height of(b) the equimolar blend.
5-30
Spherolite Morph%gy
a
Figure 5.12 Reflectance optical micrographs of the mell-crystallized dendritic structuresgrown in thin sections of the fùm al 80 oc of (a) PSECH;
5-31
•Spherulite Morph%gy
b
Figure 5.12 (cont'd) Reflectance optical micrographs of the melt-crystallized dendritic
structures grown in thin sections of the fl1m at 80 oC of (b) the equimoiar
polyenantiomer blend.
5-32
•Spherulite Morph%gy
it was reported that the pronounced same-sense curvature of the flat-on lamellae rel1ects
the tendency toward same-sense twisùng in thicker. banded spherulites. not restricted hy
fllm thickness.42 In (h) the straight branches of the equimolar blend structures are visible.
The surfaces of these dendriùc structures are depicted in the corresponding low
magnificaùon AFM images in Figures 5.13(a) and (b). Higher magnificaùon AFM images
of the apparent branch points in the dendritic structures for the opùcally pure PSECH and
the equimolar blend samples are presented in Figures 5.13(c) and (d), respectively.
Collectively, the lamellae are turning on edge at the branch point in the opùcally pure
PSECH polyenanùomer structure. However, in the equimolar blend structure the
branches appear to he composed of layers of flat lamellae, each layer measured to have a
thickness of the order of 100 Â.
The dendritic structures in Figures 5.12 and 5.13 are similar in appearance to the
crystal aggregates of the banded, a-phase of poly(vinylidene fluoride), (PVF2), grown
from very <Iilute melts of a binary blend with a non-crystallizable polymer as observed by
Khoury and Briher.43 The structures displayed radially twisting lamellae that periodically
fanned-out with a period which corresponded to that of the banded spherulites formed
from the melt under the same thermal conditions. The twisting phenomena was attributed
to unequal fold surfaces due to staggered chain folds, as described by Keith and Padden
for polyethylene,9 while the splaying feature was attributed to a periodic release of stress
build-up between stacks of lamellae. Without mention of twist sense, it was suggested
that the combined twisting-splaying motif is the underlying structure in banded spherulites.
as opposed to simply helicoidal twisting. Although the structures formed from a viscous,
amorphous heterogeneous melt and the present dendritic structures formed in thin sections
of a homogeneous melt, have apparently similar morphologies. we consider the splaying in
the poly(epichlorohydrin) structures to he indicative of an insufficient supply of
crystallizable material.
5-33
•Spherulitt Morph%gy
a
o 10 20 30 1111\
b
o 5 10 15 20 25,nll
Figure 5.13 Law magnification AFM images of the dendritic structures depicted in
Figure 5-7 of (a) the optically pure polyenantiomer PSECH and (b) theequimolar blend (images collected using constant height mode);
5-34
•Spherulite Morph%gy
c
•d
o 2 3 4 Sil'"
•Figure 5.13 (cont'd) Higber magnifieatiOII AFM images of the dcndritic structures
depicted in Figure 5-7 of (c) the optica11y pure polyenantiomcr PSEΠand(d) the cquimolar blend. (unagcs collccted using constant height mode).
5-35
•Spherll/ite Morph%gy
5.3.4 On Lamellar Twisting
Ordered arrays of identical geometrical unil~ giVt: rise to a regular. repeated
pattern upon observation at a larger level. The overall shape of the isochiral lamellae
refleclS the molecular geometry of the closely packed. same-sense helices l'rom which they
are constituted. The ordered arrays of these densely-packcd. isochiral lamellae
consequently imparts a characteristic surface appearance to the spherulites formed l'rom
the optically active macromolecule. Consider a slice of an isochiral lamella dissecting the
radial growth direction at right angles. so that it contains essentially a monolayer of chain
stems of isochiral helices. Such an arrangement can be considered as a two dimensional
array of long chiral rods of a given handedness. This arrangement can also he delined hy
an idealized plane (containing the length of the rods) of the nematic phase of a liquid
crystal. It has been shown for liquid crystalline polymers that when the rods are ail
isochiral. the free energy is lowered when neighboring planes align the long axis of thcir
respective isochiral rods slightly less than paralleI.44-4S The eITect of the backhonc chirality
of the chains in a chiral nematic phase is thus to generate the resultant cholesteric phase
in which the direction of molecular chain stem orientation rotates in a periodic, hclical
fashion. This is ilIustrated in Figure 5.14 by a theoretical illustration of the chiral ncmatic
phase created in a cellulosic suspension. By analogy to cholesteric liquid crystalline
polymers, it is conceivable then that regularly twisted. helicoidal isochiral lamellae could
result l'rom the effect of the lowering of the lateral free energies of the composite folded·
chain isochiral helices. In this way, the two possible twist directions of a lamella
composed exclusively of chiral helices of one handedness are expected to be energetically
non-degenerate. What emerges is a true chiral factor which determines the direction of
lamellar twist. It is attributed to the handedness of the constituent molecular helices,
which in turn is determined by the macromolecular backbone chirality. Indeed, the
inclination of the concerted lamellae in the low magnilication AFM images of the
spherulitic surfaces depends directly on the chiral nature of the polyenantiomers
5·36
•Spherulire Morph%gy
Figure 5.14 A theoretical illustration of a cellulosic suspension showing a chiral nematic
phase behavior.
5-37
•Splzerulire Morph%gy
comprising the isochiral spherulite. Similarly. the observation of lhe unique spherulitic
spiral sense. dependent upon the chirality of the constituent polyenantiomer. is in keeping
with the transmission of lhe efleclS of the molecular geometry lO the higher order
structures. The observed dependence of spherulite morphology on backbone chimlity has
not heen reported previously for melt-crystailize(1 synthetic polymer.~. However. it can he
related to a similar observation by Lotz et al.,46 in a study of folded-chain single crystals
and crystal aggregates of silk fibroin and ilS model enantiomer polypeptides l'rom solution.
The direction of the lamellar twist was constant along the length of the lamella and was
observed to he dictated by the chirality of the asymmetric amino acid residues in lhe
chains, Le., it had an intralamellar geometry origin. Although the pitch of the larndlar
twist in the polypeptide crystal structures is comparable to the bandwidth in banded
spherulites, the polypeptides do not have a similar bandwidlh dependence on
crystallization temperature.
Indeed, m this previous work by Lotz et al, the successive rotation of the
neighboring chain stems was compared specifically to the rotation of lhe chain axes in lhe
cholesteric liquid crystalline fonn of polY('Y-henzyl glutamate).44 They maintained thal lhe
intralamellar origin of the geometry involved the core of the lamella. They noted lhal
although the presence of surface stresses l'rom non-orthogonal chain stems cannol he
ruled-out as a contributing factor to the twist, il cannot entirely accounl for the lamellar
geometry.
In genenù, the chain stems in a lamella can maintain a non-orthogonal relationship
to the opposite fold surfaces either by canting toward the radius of the spherulite or in a
tangential direction. Keith and Padden9 have predicted that the surface stresses which
result l'rom the fonner will cause the fonnation of predominanûy helically twisted
lamellae. There are reports of banded polyester spherulites in which the chains are known
to he inclined toward the radius.47 For chain canting in the tangential direction they
predict the fonnation of helicoidally twisted lameUae, as in the case of banded
5·38
•Spherulire Morph%gy
polyethylene spherulites. Although the chain stems are successively rotated in the
tangential direction in the cholesteric liquid crystal phase analogy. this does not rcquire
that the chains he tilted with respect to the lamellar normal. It is not known whether or
not the chains are canted in poly(epichlorohydrin). However, it should he noted that the
c-axes of regularly rotating chain stems in helicoidally twisted lamellae can still maintain an
orthogonal rclationship with the lamellar fold surfaces. as shown in Figure 5.15.
Radial Growth
Figure 5.15 A schematic reprcsentation of a radiating helicoidallamella with the chain
direction maintaining an orthogonal relationship with the top and bottom
lamellar fold surfaces (taken l'rom reference 46).
5-39
•Sphu/llite Morph%gy
5.3.5 Asymmetry as a Simple Chiral Factor
Surely, beyond the contines of the planar zigzag polyethykne system there are
lamellar lWist-determining factors other than chain eanting. In his studies of PVF~.
Vaughan26 found that although the chains in the 'Y-phase spherulites are known to he tilted
with respect to the lamellar normal, the lamellae have a curved hut untwisted pmlik and
banding is not observed. It was not known whether chain canting occurs in the a-phase,
which forms banded spherulites with predominantly planar lamellae demonstr.lting an
alternating flat-on to edge-on orientation along the h-axis growth direction in the
spherulite surface replicas. Lovinger et aI.48 have demonstrated that in the a-phase of
PVF2 the dipolar chains are statistically packed so that there is a cancellation of the dipole
vectors of the unit cell while in the y-phase the lamellae are composed of polar unit cells.
Thus, macromolecular backbone polarity can affect the way chains arrang" in the unit ccli
and consequently influences the organization of unit cells in the lamellae. These effecl~
are manifesteJ in the gross morphology as banded spherulites.
The extinction of any long-range cooperativity of the lamellae in the spherulitcs in
an equimolar blend of the two enantiomers lends support to a stereoselection at the
growth front in the blend spheruIites, as we have suggested earlier.21 However, the extent
to which chains of the opposite sense polyenantiomer are excluded l'rom the lamellae
remains unresolved. The exclusion of opposite-sense helices in the equimolar blend may
be incomplete. The condensation of sorne adjacent enantiomeric chains at the growth
front would interrupt a regular lamellar twist, and in sorne instances. represent poinl~ of
helicoidai twist reversai. Indeed, Cothia49 suggested that the presence of achiral residuc.~
in twisted ~-sheets of polypeptides should deerease the tendency of the lamellae to twist,
perhaps even caneeling the effee!. The lamellae may weil be turning on their edge in the
blend spheruIite. albeit randomly and without any significant intcrlamellar long-range
oroer. giving rise to a coarse. incoherent birefringent pattern and colleetively. a relatively
5·40
•Spherulite Morph%gy
flat spheculitic surface. It is weil known that the coarse, fibrillar appearance of nonbanded
spheculites arises from the edges of the lamellae as they intersect the surface.so
ln view of the large number of crystalline species that form banded spheculites,SI
which cxtends far beyond polymeric materials, it appears that an explanation of such a
widespread phenomenon involves the very general prerequisite of a necessary level of
asymmecry. Such a gener.ù prerequisite can account for the common banded spheculitic
features among small moiecules and folded-chain structures while still recognizing the
inherent differences among these crystalline species. Apparently, the element of
asymmetry is not essential at the level of the residue, specifically. For example, in
polyethylene, a~ymmetry is introduced only at the level of the tertiary structures, Le., the
compression anisotropy at the opposite lamellar fold surfaces induced by the chain tilting
within one lamella In the case of banded spherulites of ~-polypropylene,the asymmetrical
unit appears to he present hl the level of the secondary structures in the form of pseudo
chiral helices. As for the more complex PVF2 system,26 the presence of more than one
element of asymmetry may be influencing the lamellar organization, namely, those of
macromolecular polarity and chain canting. The PVF2 case may also indicate the
importance of the strength of the asymmetric element Ultimately. nllt every asymmetric
element will have such pronounced effects on the crystalIization process.
5.4 Summary and Conclusions
In the highiV organized poly(epichlorohydrin) system the backbone chirality of the
polymer chains imposes significant restrictions on the lamellar organization within the
spherulites of both of the e.1lI1ltiomers and their equimolar blend. The observed
differences in the spherulitic surface morphologies among the polyenantiomers and their
equimolar blend support our initial suggestion that a stereoselective mechanism occurs at
5-41
•Spherulire Morph%gy
the growth front in the blend spherulite. The significant loss of coordination among the
lamellae in the equimolar blend in both the radial and circumfercntial directions of the
unrestrained spherulitic surface supports this c1aim. In contrast to the long-range in
phasing of the lamel1ae in the polyenantiomer spheruliles. the equimolar hlend spheruliles
manifest only patches of local, short-range order. Bunches of lamel1ae collectively give a
relatively flat surface. Often neighboring bunchc.~ are not in close contact but are
separated by gaps in the surface of the order of 1000 Â deep. However. the extent of the
exclusion of oppositely handed helices from the lamel1ae in the equimolar blend system is
not resolved. An incomplete stereoselection mechanism at the growth front of the
equimolar blend spherulites is suggested, where the periodic incorporation of opposite
sense chains present locations of regular lamel1ar twist interruption and/or helicoidal twist
reversaI.
The observed surface morphology of the banded, isochiral spherulites of PSECH
and PRECH is consistent with previously modeled surfaces of a spherulitic architecture
composed of an assemblage of radiating helicoidally twisted lamellae,lo No evidence was
seen to indicate that regularly spaced screw dislocations aIong the lamellae accommodatc
the twisted molecular orientation in the banded spherulites. as described in the model
developed by Bassett.6 The observed morphology favors the existence of an underlying
lamellar geometry of radiating, regularly twisted helicoidal lamel1ae. as in the model by
Keith and Padden.9 The nature of the intraIamellar origin of such a geometry, however,
appears to depend on the molecular characteristics of the poly(epichlorohydrin), It is not
known whether the chains are tilted with respect to the lamellar nonnal in opticaIly active
poly(epiclùorohydrin). However, for these polyenantiomers il would seem that the factor
which detennines the direction of twist of the lamellae is the chiraI identity of the chains.
Although stresses at the opposite fold surfaces cannot he excluded. it is suggested that the
twist of the lamellae arises from the lowering of the lateraI surface free energies of the
isochira1 helices as they condense in slightly less than parallellayers in the lamel1ar growth
5·42
Spheru/ite Morph%gy
direction. Indeed, the molecular geometry of the isochiral helices is translated beyond the
lamellar level to that of the spherulite, the surface of which manifests a spiral, the direction
bcing dependent upon the handedness of the constituent polyenantiomer. For optically
active poly(epichlorohydrin) the asymmetrical units present in the backbone of the
optically pure polyenantiomer are actively influencing the development of the overall
spherulite and their effects are transmitted to the level of the gross morphology.
5.5 References
(1) Keller, A. J. Polym. Sei. 1955,17,351.
(2) Fischer, E. W. Z. Naturforsch 1957, 12A, 753.
(3) Price, F. P. J. Polym. Sei. 1959, 39, 139.
(4) Fujiwara, Y. J. App/. Polym. Sei. 1960,4, 10.
(5) Keller, A; Sawada, S. Makromol. Chem. 1964, 74, 190.
(6) Bassett, D. C.; Hodge, A. M. Polymer 1978,/9, 469.
(7) Bassett, D. C.; Hodge, A. M. Proc. R. Soc. Land. 1981, A377, 25.
(8) Bassett, D. C.; Olley, R. H.; Al Raheil, A. M. Polymer 1988, 29, 1539.
(9) Keith, H. D.; Padden Jr., F. J. Polymer 1984, 25, 28.
(10) Lustiger, A.; LoIZ, B.; Duff, T. S. J. Polym. Sei.• Polym. Phys. Ed. 1989,27,561.
(Il) Toda, A.; Keller, A. Colloid Polym. Sei. 1993,271,328.
(12) Bassell, D. C.; Freedman, A. M. Prog. Colloid Polym.Sci. 1993, 92, 23.
(13) Snetivy, D.; Vansco, G. J. Polymer 1994, 35,461.
(14) Lotz. B.; Wittman, J. C.; Stocker, W.; Magonov, S. N.; Cantow, H.-J. Polym.
Bull. 1991,26,209.
(15) Snetivy, D.; Guillet, J. E.; Vansco, G. J. Polymer 1993, 34, 429.
(16) LOIZ, B. Phil. Trans. R. Soc. Land. A 1994, 348, 19.
5·43
•SplJerulire Morphology
(17) Bruckner, S.; Meille. S. V.; Petraccone. V.; Pirozzi. B. Pro.~. Polym. Sei. 1991.
16,361.
(18) Lotz. B.; Wittrnann, J. C. J. Poly. Sei.. Polym. Phys. Ed. 1986.24. 1541.
(19) Stocker, W.; Magonov, S. N.; Cantow. H.-J.; Wittmann. J. C.; Lotz. B.
Macromolecules 1993. 26. 5915.
(20) Li-Sheng. L.; Stupp. S. I. Macromolecules 1995. 28. 2618.
(21) Singfie1d. K. L.; Brown. G. R. Macromolecules 1995, 28. 1290.
(22) Keller. A. J. Polym. Sci. 1959.39.151.
(23) Keith. H. D.; Padden. F. 1. J. Polym. Sei. 1959.39. 101.
(24) Keith, H. D.; Padden, F. J. J. Polym. Sei. 1959,39. 123.
(25) Patel. D.; Bassett. D. C. Proc. R. Soc. Land. A 1994. 445. 577.
(26) Vaughan. A. S. J. Mater. Sei. 1993. 28. 1805.
(27) Schultz. J. M.; Kinloch. D. R. Polymer 1969. JO. 271.
(28) Bassett, D. C.; Vaughan. A. S. Polymer 1985. 26. 717.
(29) Keller. A. J. Polym. Sei. 1955, 17. 291.
(30) Hoffmann, J. D.; Lauritzen, J. J. J. Res. Natl. Bur. Suis. 1961. 65A. 297.
(31) (a) Olley, R. H.; Hodge. A. M.; Bassett, D. C. J. Polym. Sei.. Polym. Phys. Ed.
1979.17.627. (h) Bu, H. S.; Cheng. S. Z. D.; Wunderlich. B. Polymer 1988. 29.
1603. (c) Palmer. R. P.; Cobbold, A. Makromol. Chem. 1964. 74. 174. (d) Priest,
D. J. Polym. Sei. A-21971, 9. 1777.
(32) Wittrnan. J. C.; Lotz, B. J. Polym. Sci.• Polym. Phys. Ed. 1985. 23, 205.
(33) Kanig, G. Kolloid Z. 1973.251,782.
(34) see reference 16 and other papers in this series.
(35) (a) Fischer, H.; Miles, M. J.; Odell, 1. A. MacromoI. Rapid Commun. 1994. 15.
815. (h) Patnaik. S. S.; Bunning, T. J.; Adams. W. W.; Wang, 1.; Labes. M. M.
Macromolecules 1995. 28, 393.
5·44
•
•
•
Spherulite Morphology
(36) see for example, Shakesheff, K. M.: Davies, M. C.: Roberts, C. J.: Tend1er. S. J.
B.: Shard, A. G.: Domb, A. Langmuir 1994, 10,4417.
(37) Dietz, P.: Hansma, P. K.: Thn, K. 1.: Motamedi. F.; Smith. P. J. Mater. Sei. 1993,
28, 1372.
(38) Howland, R. S. How to Buy a Scanning Probe Microscope: Park Scientilïc
Instruments, 1993.
(39) (a) Geil, P. H. In Polymer Single Crystals; Mark. H. F., Immergut, E. H.• Eds;
John Wiley & Sons: New York, 1963; p. 230. (b) see p. 233 in (a). (c) see p. 247
in (a).
(40) (a) Keller, A. In Growth and Perfection of Crystals; Doremus, R. H., Roberts. B.
J., Turnbull, D., Ed.'s; Wiley: New York, 1958; p. 499. (b) Geil, P. H. J. Polym.
Sei. 1961, 51,510.
(41) Tbornton, A. W.; Predecki, P. J. Appl. Phys. 1970,41,4266.
(42) Keith, H. D.; Padden, F. J., Jr.; LOIZ, B.; Wittman, J. C. Macromolecules 1989,
22,2230.
(43) Briber, R. M.; Khoury, F. J. Polym. Sei.• Polym. Phys. Ed. 1993, 31, 1253.
(44) Robinson, C. Tetrahedron 1961, 13, 219.
(45) Grosberg, A. Y. Sov. Phys. Dokl. 1980,25,638.
(46) LOIZ, B.; Gonthier-Vassal, A.; Brack, A.; Magoshi, J. J. Mol. Biol. 1982, 156,
345.
(47) Keith. H. D. Macromolecules 1982, 15,114,122.
(48) Lovinger, A. 1.; Amundson, K. R. Polym. Prepr. (Am. Chem. Soc.• Div. Polym.
Chem.) 1993,31, 784.
(49) Cothia, C. J. Mol. Biol. 1973, 75.295.
(50) Claver. Jr., G. C.; Buchdahl. R.; Miller, R. L. J. Polym. Sei. 1956, 20, 202.
5-45
•Spherulire Morph%gy
(51) (a) Ryschenkow. G.: Faivre. G. J. Crysr. Gro",rh 1988. 87. 221. (h) Bisault. J.:
Ryschenkow. G.: Faivre. G. J. Crysr. Gro",rh 1991. 110. 889. (c) Lagassc. R. R.
J. Crysr. Growrh 1994. 140.370. (d) sec refcrencc 39.
5-46
Conclusions. Contributions & Continued Research.
CONCLUSIONS,CONTRIBUTIONS
TO ORIGINAL KNOWLEDGE,& IDEAS FOR CONTINUED
RESEARCH
ChapterSix
6.1 Conclusions &Contributions to Original Knowledge
The flfst reported synthesis and characterization of the optically pure
polyenantiomers of poly(epichlorohydrin) (PECH) in Chapter Two demonstrates that the
polymerization of epichlorohydrin monomer using the triethylalwninwn:water catalyst
proceeds with a ring-opening attack exclusively at the ~-earbon position. Stereo
irregularities and regio-irregularities are absent from the polyenantiomer made from either
optically pure monomer. The separated acetone-soluble fraction of the polymer prepared
from the racemic monomer in similar fashion, is irregular by nature of its stereo-
irregularities only. The quantitative dechlorination of PECH generates the high molecular
weight poly(propylene oxide) POO, without otherwise altering the asymmetric centers.
The dechlorination of the optically pure polyenantiomers of PECH reported in Chapter
Two is a convenient method of attaining both optically pure forms ot high molecular
weight POO, which have not been reported in the literature.
6-1
•Conclusions. Contributions & Continlled Research.
The availability of the individual polyenantiomers in thcir pure forms permil~ the
ready investigation of the properues of the equimolar polyenantiomer hlend in the case of
PECH or PPrO. The need for speculation on the properties of a truc polyenantiomer
equimolar blend, based on those observed l'rom a proposed "statistical mixture" of Rand S
chains resulting l'rom a stereoselective polymerization of the racemic monomer. i~
a1leviated. From a basic investigation of the thermal properties of these optically active
polyethers, il is concluded that neither the equimolar blend of the PECH polyenantiomers.
nor the PPrO polyenantiomers form a stereocomplex l'rom the melt For each polyether
system, the different forms of the polymer develop crystallites of varying metastahility
upon crystallization l'rom the melt but contain the same hasic crystal structure, as
evidenced by a common equilibrium temperature. However, the stereoblock form of
PECH, specifically, possesses a certain degree of disorder, which the polyenantiomers and
their equimolar blend do not
A more thorough investigation of the thermal properties of PECH in Chapter
Three indicates a slightly slower rate of crystallization for the polyenantiomer blend. and a
dramatically reduced rate of crystallization of the stereoblock, compared to the
polyenantiomers The differences among the polymers can only stem from their relative
structural differences which, specifically, are the presence of stereo defect sites in the
stereoblock polymer, and the presence of the opposite-sense polyenantiomer in the case of
the blend. Crystallites of the stereoblock are fewer, smaller, and less stable against me1ting
!han those of the optically active polyenantiomers or the equimolar polyenantiomer blend;
long residence limes at high temperatures improve the thermal stability.
Both PECH and POO polyethers display characteristic multiple melting
endotherms in the thermograms of isothermally crystallized samples. In each case, the two
lowest melting peaks are representative of the initial distribution of crystallites. The
lowest temperature peak is due to the melting of subsidiary lamellae, formed during
6-2
•Conclusions. Contributions & Continued Research.
secondary crystallization; the middle temperature peak represents the melting of the
initially formed dominant lamellae; the high temperature peak is due to the melting of
material recrystallized material during the thermal analysis heating scan. The melting of a
sample subsequent to crystallization at large undercoolings is dominated by
recrystallization processes. The observed melting thermogram of a sample crystallized at
low undercoolings. however, is characteristic of the melting of the originally present
dominant and subsidiary lamellae.
The nature of the investigation in Chapter Four is unique in heing the flfSt
puhlished comparison of the radial growth rates of two optically pure polyenantiomers,
their equimolar blend, and a stereoblock form of the same polymer. By virtue of the
structural relationship among the polymers in question, the observed differences in
crystallization kinetics among them can only he accounted for in terms of a stereo-related
process. Specifically, it is concluded that there is an inherent preference of stereospecific
sequences, which can constitute either part or ail of a chain, to self-assemble rather than to
co-crystallize with those of the opposite sense. The implications of the tendency for
stereosegregation of the polymer chains at the growth front, are a reduced rate of
crystallization with increasing amount of opposite-sense sequences. As segments of the
same polymer chain, the added effeet of intimacy of the opposite-sense sequences
suppresses the rate of crystallization even more.
The reduction in the spherulite radial growth rates of the stereoblock relative to
the polyenantiomer, for both PECH and PPrO, can he accounted for in terms of the role of
the stereo defeet sites in the stereoblock during crystallizationlsegregation. In the case of
PECH, the increase in the nucleation term of the i-PRSECH arises from a large fold
surface free energy of the crystallites. The stereo defeet sites in the i-PRSECH polymer
have a strong influence on the nature of the lamellar fold surfaces, causing the formation
of loose folds and the increased amounts of moleeular cilla, relative to the polyenantiomer.
6·3
•
•
•
Concillsions. Contributions & Continlleti Resellrch.
In addition. the rejected opposite-sense stereosequenœs of the stereohlock may contrihute
ta the amorphous regions as tie molecules helWeen lamel1ae. In the case of PPrO. the
mechanism of accommodation of an opposite-sense stereosequence occurs with much less
disruption, both to the rates of radial growth and to the resulting morphology. For hoth
PECH and PPrO. branching and thus the formation of subsidiary lamel1ae hccome a factor
ooly at high crystallization temperalUres, as evidenced by the observed distortion of the
birefringent extinction pattern in the banded spherulites.
By a comparison of the isolaetie, optically pure, regio-regular polyenantiomers of
each polyether, il ean he said that PECH manifests a slower tendency to crystalli7.e than
PPrO. Thus in slUdying the factors which govern crystallization, through the prohing of
the effects of opposite-sense stereosequences, interference has more readily taken place in
the PECH polymers than in the more readily crystallizable PPrO polymers.
The work in Chapter Five marks the fust report of the direct observation, hy
atomic force microscopy (AFM), of spherulite spiral sense in isochiral, banded spherulites.
The visual persuasiveness of the direcdy observed spherulitic chiral surface patterns of the
PECH polymers substantiates the conclusions put forth in the preceding chapters. The
isochiral. banded spherulite architecture is struclured upon radiating helicoidally lWisling
lamellae. The chiral faClOr of the coherendy lWisting lameUae is the isochiral sense of the
asymmetric carbans of the constituent polyenantiomer chains. The drive 10 lwisl can arise
from a lowering of the lateral surface free energies which result from the configuration of
the isochiral chains within a twisting lamella. The combination of opposite-sense
stereosequences, either as a stereoblock or as an equimolar polyenantiomer blend, causes
an extinction of the long-range cooperativity of the lamellae by interrupting the regular
helicoidal lamellar twist, and effects the formation of non-banded spherulites. The
asymmetrical units in the optically pure polyenantiomer chain actively influence the
6-4
Conclusions. Contributions & Continued Research.
devclopment of the overall spherulite and their effecl~ are transmitted to the level of the
gross morphology.
The study presented in Chapter Five is unique in employing the property of optical
activity as an analytical tool in the elucidation of spherulite morphology. It is a signilicant
contrihution to the Iiterature on banding in spherulites. The study completes a so-far
elusive symmetrical relation between behavior in monomerics and in polymers, viz. when
present, structural asymmetry of the kind underlying optical activity also governs
appearance and handedness of twisting orientation in spherulites.
6.2 Ideas for Continued Research
The polyethers used in this thesis contain three backbone atoms per monomer
repeat unit. As outlined in Chapter One, the odd number of backbone atoms restricts the
pendent group to alternating sides of the chain in an isotactic configuration. It would be
very interesting to synthesize the corresponding syndiotactic PECH polymer. AlI of the
pendent CHzCI groups would be situated on one side of the polymer chain. This
configuration would result in an increased Van der Waals repulsion between the
neighboring chlorine atoms. As the chlorine atoms shift in space finding a lower energy
conformation, the effect on the overall chain would be ta form a helix witil a definite sense
and a pitch which is altered relative to the one for the isotactic chain. Sti"reocomplexation
may be a factor in the crystallization of the two opposite-sense syndiotactic
polyenantiomers. Similarly, these polymers could be dechlorinated to form the
corresponding syndiotactic POO polymers.
The dechlorination of PECH to POO could be used as a surface treattnent
technique in a potentially very elegant lamellar surface analysis. The analysis could be
6-5
•Conclusions. Contributions & Continued Reuorch.
pcrfonned in a bulk fashion on a highly crystalline Hlm. or the suhject lamellae could he
those available in a sample such as the dendritic structures on silicon wafcr supports in
Chapter Five. The choice of sample preparation would detennine the Hnal method of
surface characterization. Essentially. the lamellae of PECH would he treated with a
solution of LiAlH4 in ether (now available l'rom Aldrich Chemicals). in which solvent thcre
is no swelling of the PECH. Only the amorphous parts of the sample. Le.. the fold
surfaces of the lamellae. would undergo the dechlorination reaction. Thus the large loops
and molecular cilla which contribute 10 the surface disorder would he convened to PPrO.
The lamellae would then he washed. liltered. and the liltratc preserved. The
dechlorinated. soluble PPrO surface molecules. now severed from the crystallite and in the
liltrate. would he analyzed using solution nuclear magnetic resonance (NMR)
spectroscopy. The surface modified lamellae would he investigated (hefore and after
surface treatrnent) by perfonning solid state NMR on the rùm sample. A~ the NMR
technique is a non-destructive one. the resulting lilm could also he suhject to DSC analysL~
to investigate whether reorganization of dominant lamellae has been restricted hy the
removal of the loose ends which would nonnally he annealed into the polymer crystallite.
Ideally. the relative amounts of disorder at the fold surfaces of the lamellae of the
stereoblock and the optically pure polyenantiomer could he detennined. Moreover. if the
resulting POO surface molecules are subject to analysis by GPC. an estimate of the size of
a fold surface loops or rejected chain ends could be detennined.
The dendritic structures which develop in the melt-deprived regions of the PECH
samples shown in Chapter Five. could he subjected to a lamellar thickness study using
AFM. Specifically. the step heights of the lamellar islands could he measured on the same
sample as a function of crystallization time at a single crystallization temperature. or as a
function of isothermal crystallization temperature. This technique could he used as an
6-6
•Conclusions, Contributions & Continued Research.
alternative ta the small angle X-ray scattering (SAXS) technique nonnally employed to
pcrfonn larnellar long period measurements.
A1though the PPrO polymers have received limited attention in the present work,
serving mainly to investigate the added effect of substitution of the Cl atom on PECH with
H, il has becorne obvious that a well-fractionated optically active PPrO system rnerits a re
exarnination of the growth rate kinetics with an evaluation of the high temperature data as
weil as a full investigation into the many interesting phenomena which it manifests.
il is evident that the necessary AFM investigations should he performed on the
PPrO polymers. What needs to be determined is whether a similar spherulite chiral surface
pattern is also characteristic of the optically pure PPrO polyenantiomers. If the banded
spherulites of PRPrO and PSPrO are distinguishable as in the case of PECH then it may he
possible to identify regions of PRPrO and PSPrO in the sarne banded equimolar blend
spherulite using AFM. The idea of different asymmetty elements which can cooperate or
compete would he a theme in the study of the banded spherulite morphology of PPrO.
That is, unIike PECH, the relatively high flexibility of the dechlorinated polyether may
hinder or enhance the influence of the backbone chirality on the resulting morphology of
the helix, larnellae, and fmally spherulite. Figure 6.1 (a) contains the atomic force
micrograph of a prelùninary scan of a sarnple of PSPrO. The relatively larger bands of the
PPrO polymers, compared to PECH, result in larger areas of flat and edge-on larnellar
orientations. There is thus a greater possibility of attaining the image of a lateral surface
(an edge) of the larnella. With high resolution, this would permit tht imaging of the
molecular chains along their length. In Figure 6.I(b), the polarized light photomicrograph
of i-PRSPrO shows approximately one quadrant of a spherulite. In this section, however,
the !ibrils of the spherulite appear to he twisting. Fibrils are a higher level morphological
form composed of larnellae. In this spherulite, the bands are present but are difficult to sec
6-7
•
Conclusions. Conrribu1ions & Conrinued Research.
in the photomicrograph. due to the fact that the focus is on the lihrils. This is very
intriguing. and may imply that in PPrO. the transmission of chimlity may he to the level of
the fibrils. The surface analysis of such a sample hy AFM would easily determine the
direction of the twist.
The direction of the apparent surface spiral in this example spherulite is Idt
handed. the same as that that found for PSECH. although it must he noted that whether
this spherulite is representative of ail PSPrO spherulites is entirely unknown at this point.
However, the digitized video images of the dendritic structures of this polyenantiomer in
Figure 6.2, show pinwheel-Iilœ dominant lamellae tuming in the same direction to that
observed for PSECH. It is noteworthy to add howe\'er, that the PSPrO polymer in
Figures 6.1 and 6.2 is derived l'rom the PRECH polyenantiomer. Recall. that the ahsGlute
configuration of the asymmetric center changes upon substitution of the chlorine atom
with hydrogen, in the reduction of PECH to PPIÛ.
The mcchanism of accommodation ur the opposite-sense stereosequence in the i
PRSPrO stereoblock is .ilio a very interesting problem to address. Again. the
dechlorination reaction of PECH may he useful: Reduction of PECH using LiAlD4 yield~
the methyl-deuteralCd PPrO polymer. With only one deuterium on the pendent methyI.
the modification is subtle and should not play a role in the crystallization kinetics of the
polymer, but will provide a tag which is readily detectable by Fr-IR (Fourier transform
Infrared) spectroscopy. Monitoring the isothermal crystallization process starting l'rom
the lime of the melt, may provide sorne insight into the details of the segregation process,
if any, in the equimolar polyenantiomer blend.
Figure 6.3 contains the reflectancc light photomicrographs of a sample of PRECH
which was typically prepared for 3l'.lllY5is by AFM according to the method outlined in the
Chapter Five. The sample was phntographed after the removal of the top silicon wafer by
6-8
•
•
•
Conclusions. Contributions & Continued Research.
immersing the sample in Iiquid nitrogen. These resulting fracture patterns were routinely
ohserved, which trace the band lines of the spherulites. It was found that the uneven
surface of the resulting fractured spherulite is not sufficiently flat to allow for analysis by
AFM. However, it might he very useful to apply the technique of scanning electron
microscopy to these samples. It is not known whether the cracks arise due to the adhesion
of the polymer to the silicon wafer or due to polymer shrinkage which would occur when
the sample is immersed in the Iiquid nitrogen, or both. In any case, the resulting images of
these samples and those non-isothermally crystallized, i.e., with a continuously changing
hand period, may serve to provide dues as to the location of the weak points of a
helicoidally twisted lamella.
6-9
•
Figure 6.1
a
Cane/usions, Contributions & Conrinutd Rtuarch.
2O~m
(a) Low magnification AFM image of PRPrOscanned in constant height
mode;
6-10
Cone/usions. Conrributions & Continued Research.
100 J.1ID
Figure 6.1 (cont'd) (b) Polarized light photomicrograph of i-PRSPrO taken with a
114 À wave plate.
6-11
•Conclusions. Contriburions & Continued Research.
b
Figure 6.2 Digitized video images of dendritic structures of PRPrO formed in very
thin sections of the melL
6-12
Conclusions. Contributions & Continued Research.
a
b
Figure 6.3 Reflectance light photomicrographs of PRECH spherulites. showing thecircumferential fractures. Details in the text
6·13
