Luận án tiến sĩ: Fundamental and applied research enabled by polymer nanolayer coextrusion technology

The transverse direction T is vertical and the film normal direction 1.8 Two dimensional transmission diffraction patterns of the PP control and films with 460 nm, 108 nm, 65 nm, and 10

FUNDAMENTAL AND APPLIED RESEARCH ENABLED

BY POLYMER NANOLAYER COEXTRUSION TECHNOLOGY

Dissertation Advisers: Dr Anne Hiltner and Dr Eric Baer

Department of Macromolecular Science and Engineering CASE WESTERN RESERVE UNIVERSITY

May 2007

3241778 2007

UMI Microform Copyright

All rights reserved This microform edition is protected against unauthorized copying under Title 17, United States Code.

ProQuest Information and Learning Company

300 North Zeeb Road P.O Box 1346 Ann Arbor, MI 48106-1346

by ProQuest Information and Learning Company

SCHOOL OF GRADUATE STUDIES

We hereby approve the dissertation of

*We also certify that written approval has been obtained for any

proprietary material contained therein

To sister (金鑫) and brother (金大祥)

To husband (傅朝阳)

iii

Page

LIST OF TABLES vi

LIST OF FIGURES vii

ACKNOWLEDGEMENTS xv

ABSTRACT xvi

CHAPTER 1 STRUCTURE OF POLYPROPYLENE CRYSTALLIZED IN CONFINED NANOLAYERS 1

2 FORMATION AND TRANSFORMATION OF SMECTIC POLYPROPYLENE NANOPARTICLES 37

3 FRACTIONATED CRYSTALLIZATION OF POLYPROPYLENE PARTICLES PRODUCED BY NANOLAYER BREAKUP 62

4 EFFECT OF AN ORGANIC DICARBOXYLIC ACID SALT ON FRACTIONATED CRYSTALLIZATION OF POLYPROPYLENE PARTICLES 96

5 EFFECT OF A SORBITOL NUCLEATING AGENT ON FRACTIONATED CRYSTALLIZATION OF POLYPROPYLENE PARTICLES 134

6 BIO-INSPIRED POLYMER GRIN LENSES BY NANOLAYER FORCED-ASSEMBLY 162

iv

APPENDIX

A ACHIEVE HIGH QUALITY GRIN LENSES 203

B SUPERIOR GRIN LENS 214 BIBLIOGRAPHY 226

v

Table Page

1.1 Comparison of PP layer thickness measured from AFM images

with PP layer thickness calculated from processing parameters 22

1.2 Thermal properties of polypropylene in films 23

1.3 Long period measured from the SAXS pattern 24

2.1 Thermal analysis of PP nanoparticles 51

3.1 Thermal analysis of PP nanoparticles 79

3.2 Melting Enthalpies of PP particles from different PP nanolayers 80

4.1 Crystallization enthalpies of PP nanolayers 114

4.2 Crystallization enthalpies of PP particles from 12 nm layers 115

4.3 Melting enthalpies of PP particles from 12 nm layers 116

4.4 Crystallization enthalpies of PP particles from 20 nm layers 117

4.5 Crystallization enthalpies of PP particles from 40 nm layers 118

4.6 Crystallization enthalpies of PP particles from 200 nm layers 119

5.1 Melting enthalpies of PP nanolayers 148

5.2 Crystallization enthalpies of PP particles 149

5.3 Melting enthalpies of PP particles from 12 nm layers 150

vi

Figure Page

1.1 AFM phase images of cross-sections from layered PP/PS films: (a)

PP/PS 90/10 250 μm thick film; (b) PP/PS 20/80 250 μm thick

film; (c) PP/PS 10/90 250 μm thick film; and (d) PP/PS 10/90 40

μm thick film 25

1.2 Polarized light micrographs: (a) PP control film; (b) film with 460

nm PP layers; (c) film with 108 nm PP layers; (d) film with 65 nm

PP layers; and (e) film with 10 nm PP layers 26

1.3 AFM height images: (a) a typical PP spherulite at a free film

surface; (b) a discoid in a 460 nm PP layer; (c) a discoid in a 108

nm PP layer; (d) a discoid in a 65nm PP layer; and (e) a smooth 10

nm PP layer 27

1.4 AFM phase images showing lamellar morphology: (a) PP control

film; (b) a 460 nm PP layer; (c) a 108 nm PP layer; (d) a 65 nm PP 28

1.5 SAXS patterns in the flow direction (F): (a) PP control film; (b)

film with 460 nm PP layers; (c) film with 108 nm layers; (d) film

with 65 nm PP layers; and (e) film with 10 nm PP layers The

transverse direction (T) is vertical and the film normal direction

1.8 Two dimensional transmission diffraction patterns of the PP

control and films with 460 nm, 108 nm, 65 nm, and 10 nm PP

layers in the normal (N), transverse (T), and flow (F) directions 32

1.9 Pole figures of α crystals in films with 460 nm PP layers: (a)

normals to (110) planes; (b) normals to (040) planes; and (c)

normals to ( 113) planes 33

1.10 Schematic showing an edge on radial or mother lamella that

nucleated at the interface with (010) planes flat on The cross

hatched lamella formed by secondary nucleation on the (010)

plane The width of the radial lamella and the amount of

cross-vii

nucleates at the interface with (110) planes flat on 34

2.1 AFM phase images of the coextruded nanolayer assembly with 257

layers and PP/PS 10/90 (vol/vol) composition: (a) The

cross-section of the layered assembly with arrows identifying the very

thin PP layers; and (b) the surface of a PP nanolayer after it was

exposed by peeling the layers apart 52

2.2 The WAXD curves of the coextruded PP/PS assembly, the PS

control, and the PP nanolayers The latter was obtained by

weighted subtraction of the PS contribution 53

2.3 Heating thermograms of the coextruded PP/PS assembly, the PS

control, and the PP nanolayers The latter was obtained by

weighted subtraction of the PS contribution 54

2.4 The PP particles formed by nanolayer beakup: (a) AFM phase

image of the film cross-section; and (b) SEM image of the isolated

particles 55

2.5 Thermograms of the PP/PS assembly, the PS control, and the PP

nanoparticles obtained by weighted subtraction: (a) Cooling from

230 ºC at 10 ºC min -1; and (b) subsequent heating at 40 ºC min -1 56

2.6 WAXD curves of the coextruded PP/PS assembly after heating to

230 ºC for 10 min, the PS control, and the PP nanoparticles The

latter was obtained by weighted subtraction of the PS contribution 57

2.7 Effect of annealing at various temperatures for 10 min on the

WAXD pattern: (a) Before subtraction of the PS contribution; and

(b) after subtraction of the PS contribution 58

2.8 Effect of annealing at various temperatures for 10 min on the

heating thermogram of PP nanoparticles The PS contribution has

been subtracted 59

3.1 AFM phase images of the cross-sections from coextruded

nanolayer assemblies: (a) 75 μm film with 12 nm PP layers; (b)

125 μm film with 20 nm PP layers; (c) 250 μm film with 40 nm PP

layers; and (d) 1.25mm film with 200 nm PP layers 81

3.2 OM images of the particles from various layer thicknesses: (a) 12

nm layers; (b) 20 nm layers; (c) 40 nm layers; and (d) 200 nm

layers 82

viii

3.4 Particle size distributions: (a) Submicron particles from 12 nm

layers taken from AFM images; (b) large particles from 12 nm

layers taken from OM images; (c) submicron particles from 20 nm

layers taken from AFM images; and (d) large particles from 20 nm

layers taken from OM images 84

3.5 Large particle size distribution from OM images: (a) particles from

40 nm layers; and (b) particles from 200 nm layers 85

3.6 Particle volume distributions: (a) from 12 nm layers; (b) from 20

nm layers; (c) from 40 nm layers; and (d) from 200 nm layers 86 3.7 Schematic of the envisaged layer breakup proces 87

3.8 Cooling thermograms of the PP/PS assembly after heating to

230 ºC for various PP layer thicknesses The PS contribution was

subtracted 88

3.9 Heating thermograms of the PP particles from various PP layer

thicknesses after subtraction of the PS contribution 89 3.10 Dependence of crystallization enthalpy on initial layer thicknesses 90

3.11 The WAXD curves of the PP particles from various PP layer

thicknesses: (a) before subtraction of the PS contribution; and (b)

after subtraction of the PS contribution 91

3.12 Effect of breakup time and temperature on the crystallization

thermogram of PP particles from 100 nm layers: (a) Breakup time

at 230 ºC; and (b) Breakup temperature for 3 min 92

3.13 AFM phase images of the film cross section after breakup at

different temperatures: (a) 3 min at 250 ºC; and (b) 3 min at 350

ºC 93

4.1 AFM images of PP with 2 % HPN after the microtomed surface

was etched to remove the HPN particles 120

4.2 AFM phase images of the cross-sections from coextruded

nanolayer assemblies: (a) 75 μm film with 12 nm PP layers; (a)

75 μm film with 12 nm PP layers containing 0.6 % HPN; (c)

125 μm film with 20 nm PP layers; (d) 125 μm film with 20 nm PP

layers containing 0.6 % HPN; (e) 250 μm film with 40 nm PP

layers; and (f) 250 μm film with 40 nm PP layers with 0.6 % HPN 121

ix

4.4 OM images of the particles from: (a) 12 nm layers; (b) 12 nm

layers with 1.0 % HPN; (c) 40 nm layers; (d) 40 nm layers with

1.0 % HPN; (e) 200 nm layers; and (f) 200 nm layers with 1.0 %

HPN 123

4.5 AFM images of the particles from: (a) 12 nm layers; and (b) 12 nm

layers with 1.0 % HPN 124

4.6 Particle size distributions from 12 nm layers: (a) Submicron PP

particles from AFM images; (b) large PP particles from OM; (c)

submicron PP particles with 1.0 % HPN from AFM; and (d) large

PP particles with 1.0 % HPN from OM 125

4.7 Large particle size distributions from OM images from: (a) 40 nm

PP layers; (b) 40 nm PP layers with 1.0 % HPN; (c) 200 nm PP

layers; and (d) 200 nm layers with 1.0 % HPN 126

4.8 Particle volume distributions: (a) From 12 nm layers; (b) from

12 nm layers with 1 % HPN; (c) from 40 nm layers; (d) from

40 nm with 1 % HPN; (e) from 200 nm layers; and (f) from

200 nm layers with 1 % HPN 127

4.9 Thermograms of the PP/PS assembly with 12 nm PP layers

containing various concentrations of HPN after heating to 230 ºC:

(a) Cooling thermograms; and (b) subsequent heating

thermograms The PS contribution was subtracted 128

4.10 The WAXD curves of the PP particles from 12 nm PP layers with

various concentrations of HPN: (a) before subtraction of the PS

contribution; and (b) after subtraction of the PS contribution 129

4.11 Cooling thermograms of the PP/PS assembly with various

concentrations of HPN in the PP layer after heating to 230 ºC: (a)

20 nm layers; (b) 40 nm layers; and (c) 200 nm layers The PS

contribution was subtracted 130

4.12 Comparison of the crystallization thermograms of particles from

12, 20, 40 and 200 nm layers with 1 % HPN 131

5.1 AFM phase images of the cross-sections from coextruded

nanolayer assemblies: (a) 75 μm film with 12 nm PP layers; and

(b) 75 μm film with 12 nm PP layers containing 1.0 % MD 151

5.2 Heating thermograms of the coextruded nanolayer assemblies with

various concentrations of MD 152

x

5.4 AFM images of particles from 12 nm layers: (a) 0 % MD; (b)

0.3 % MD; (c) 1 % MD and (d) the PS control with 2.0 % MD 154

5.5 Particle size distributions from 12 nm layers: (a) 0 % MD; (b)

0.3 % MD; and (c) 1.0 % MD 155

5.6 Thermograms of the PP/PS assembly with 12 nm PP layers

containing various concentrations of MD after heating to 230 ºC:

(a) Cooling thermograms; and (b) subsequent heating

thermograms The PS contribution was subtracted 156

5.7 The WAXD curves of the PP particles from 12 nm PP layers with

various concentrations of MD: (a) before subtraction of the PS

contribution; and (b) after subtraction of the PS contribution 157 5.8 Dependence of crystallization enthalpy on MD concentration 158

5.9 Schematic phase diagram for the binary system in the low MD

concentration region 159

6.1 The structure of our synthetic gradient index (GRIN) lens is

inspired by the layering found in the human lens.*

Forced-assembly is used to create transparent films with thousands of

alternating nanolayers of two polymers where the relative

concentration of the polymers specifies the refractive index The

films are assembled into a GRIN material A method of molding

and cutting the layered GRIN material introduces radial

components to the index gradient and forms the biomimetic lens

*Reference 33, by permission 172

6.2 The schematic drawings illustrate the process whereby transparent

nanolayer films are stacked into a GRIN sheet and subsequently

shaped into a GRIN lens The fabricated GRIN lens contains tens

of thousands of polymer layers, comparable to the number of

protein layers in biological lenses The flexibility of the process

lies in the wide variety of index gradients that are possible in the

GRIN sheet, and in the variety of shapes that can be molded 173

6.3 The plots show the refractive index of individual nanolayer films

that are used to construct the GRIN sheets Forced-assembly

creates films with thousands of alternating layers of two polymers

with individual layers as thin as 5 nm The relative thicknesses of

the constituent layers, and hence the film composition, determines

the refractive index 174

xi

cylindrical molds are used to create a flat cylindrical GRIN lens

Alternatively, spherical molds are used to shape a radial GRIN

lens 175

6.5 The inverse dependence of focal length on lens thickness is shown

for flat cylindrical GRIN lenses (a) and flat radial GRIN lenses (b)

The lens performance compares well with the design prediction,

which is based on the linear index distribution in the GRIN sheet

and the radius of curvature in the lens molds 176

6.6 The measured refractive index profile across the plano-surface of

the plano-convex lens is compared with the parabolic design The

refractive index range is determined by the constituent polymers,

and is larger for the SAN17/(PVDF blend) lens than for the

SAN17/PMMA lens 177

7.1 AFM phase images show three PC/PMMA nanolayer films with

compositions 30/70, 50/50, and 70/30 The light layers are PC and

the dark layers are PMMA 192

7.2 Schematic drawings illustrating the process whereby a

consolidated GRIN stack of nanolayer films is shaped and polished

into a plano-convex GRIN lens 193

7.3 The checkerboard pattern is magnified through a 20 mm

plano-convex SAN17/PMMA GRIN lens that contains tens of thousands

of polymer layers 194

7.4 The shaped GRIN sheet is part of a large imaginary GRIN sphere:

(a) The parabolic radial GRIN distribution of the imaginary GRIN

sphere; (b) the almost linear GRIN distribution in the region of the

fabricated GRIN lens; and (c) the parabolic refractive index

gradient on planes perpendicular to the optical axis at a distance z

from the center of the imaginary GRIN sphere 195

7.5 The refractive index gradient on the plano surface was obtained by

measuring the composition at various positions with ATR-FTIR

microspectroscopy: (a) A series of spectra collected across the

diameter of the plano-surface of a SAN17/PMMA GRIN lens; (b)

comparison of the plano refractive index gradient from the

composition with the parabolic refractive index gradient predicted

by the lens design at various positions through the thickness of the

lens; and (c) comparison of the refractive index gradient from the

composition with the linear refractive index gradient predicted by

the lens design along the optical axis 196

xii

The GRIN distribution on the plano surface of a SAN17/(PVDF

blend) GRIN lens; and (b) the focal length measured at various

positions compared with the focal length of a non-GRIN PVDF

blend lens with the same shape as the GRIN lens 197

7.7 The flexibility of the bio-inspired approach lies in the wide variety

of index gradients that are possible in the GRIN sheet: (a) The

prescribed index distribution of the GRIN sheet (solid line) for (b)

the target plano distribution of a 6 mm plano-convex lens The

sixty 50 μm thick SAN17/PMMA films that were assembled into a

GRIN sheet are identified by solid points in (a), and the resulting

plano GRIN distribution after the sheet was shaped between a

convex mold with R2 = 4.74 mm and a concave mold with R1 =

6.20 mm and polished is described by the solid points in (b) (c)

The prescribed index distribution of the GRIN sheet (solid line) for

(d) the target plano distribution of a 3 mm plano-convex lens The

139 13 μm thick SAN17/PMMA films that were assembled into a

GRIN sheet are identified by solid points in (c), and the resulting

plano GRIN distribution after the sheet was shaped between a

convex mold with R2 = 6.20 mm and a concave mold with R1 =

7.06 mm and polished is described by the solid points in (d) 198 A.1 Experimental setup for transmission and scattering measurement 208

A.2 Surface roughness of nanolayer films: (a) extrusion direction; (b)

transverse direction 209 A.3 The plot of transmission versus the transmitted half angle 210

A.4 The Placido-cone topography of the GRIN lens convex surface,

which has a target surface radius of 23.54mm 211

A.5 The images taken from a GRIN lens and a PMMA lens with same

geometry: (a) 3-bar pattern from a PMMA lens; (b) USAF pattern

from a PMMA lens; (c) 3-bar patter from a GRIN lens; (d) USAF

pattern from a GRIN lens 212

B.1 Linear refractive index distribution in the GRIN lens along z

direction 219

B.2 Theoretical lens resolution comparing the spherical glass lens with

the superior GRIN lens 220

B.3 Shack-Hartmann wave front pattern: (a) the glass lens; (b) the

GRIN lens 221

xiii

B.5 Cross section of the focal spot image: (a) direct profile; (b)

integrated intensity 223

B.6 Compare the transmitted power between pin holes of different slit

widths 224 B.7 Images of USAF pattern: (a) the glass lens; (b) the GRIN lens 225

xiv

The generous financial support of the National Science Foundation (NSF) and the Defense Advanced Research Projects Agency (DARPA) is gratefully acknowledged

The technical support of Polish Academic of Science and Naval Research Laboratory is gratefully appreciated

I would like to thank my advisors, Dr Anne Hiltner and Dr Eric Baer, for

numerous inspirations and constant challenges to my limits

I would like to thank Andy Chang and Huiwen Tai for great mentoring and guidance

Thanks to everyone in the research group for your support and friendship In special thanks to Vickey Ronesi, Teresa Bernal, Ben Poon, Jiong Yu, Elizabeth Christenson, Peter Dias, Akshay Kamdar and Mike Ponting

I would like to express my appreciation for Chaoyang Fu, my husband Thank you for your unconditional support and encouragement

Thanks to my parents, brother and sister, to whom I can always turn for help, support and love under any condition

xv

Polymer Nanolayer Coextrusion Technology

Abstract

by

YI JIN Polypropylene (PP) and polystyrene (PS) nanolayer films were prepared The crystal structure of extremely thin PP layers confined between PS layers was studied Changes in structure were observed as the PP layer thickness decreased to the nanoscale

A dispersion of isotactic polypropylene (PP) particles was produced by driven breakup of PP nanolayers Particle size analysis indicated that breakup of PP microlayers produced a bimodal particle size distribution A population of submicron particles formed due to the Rayleigh instability, and a second population of large particles formed by relaxation Breakup of 12 nm layers resulted in primarily submicron particles, which crystallized into smectic form by homogeneous nucleation at 40 °C The fraction of

interfacial-PP as submicron particles dropped dramatically as the layer thickness increased to 40 nm Fractionated crystallization gave rise to four crystallization exotherms at higher temperatures, which represented fractionated crystallization of the large micron-sized particles in the PP α-form The effect of a particulate nucleating agent on crystallization

of polypropylene (PP) in particles was found to be vastly different the effect of a sorbitol nucleating agent

A new class of hierarchically structured polymer optical materials was demonstrated that possess an internal refractive index gradient The structure of the polymer material is

xvi

refractive index distribution can be achieved with this flexible technology within the refractive index range of available coextrudable optical materials An important application demonstrated for these materials is the construction of a bio-inspired GRIN lens

Lenses with gradients in both the radial and axial directions were fabricated Materials with gradients with Δn ~ 0.17 have been made and larger index ranges are possible

Experimentally, these lenses were shown to have the parabolic radial and nearly linear axial refractive index distribution expected from the design, similar to a section of the spherical parabolic distribution found in the lens of an octopus Comparing to a conventional spherical lens, GRIN lens exhibited improved on-axis spherical aberrations In the end, a carefully designed lens was shown to have an on-axis focal ability exceeding any off-the-shelf spherical lenses

xvii

Studies of 2-dimensional crystallization usually employ thin films with a free surface Crystallization under conditions of spatial confinement is more difficult to access experimentally In one example, the crystal structure of dispersed PVDF particles, which were crystallized in the presence of a confining PMMA matrix, changed from the more stable γ-form in larger particles to the metastable β-form in 50 nm particles.6

The minute quantity of material in ultrathin films challenges conventional techniques of polymer structure - property characterization In contrast to the well-known concept of self-assembly,7 layer-multiplying coextrusion uses forced-assembly to create thousands of alternating layers of two polymers.8 Each layer can be less than

10 nm in thickness.9,10 Although the amount of material in a single layer is very small, the properties of the layer are multiplied many-fold by the number of identical layers in the assembly This permits the use of conventional methods of polymer analysis for probing size-scale-dependent properties as the thickness of a confined layer approaches the nanoscale Assemblies of 4000 nanolayers were used to study the highly localized mixing that occurs at the interface when two immiscible glassy polymers are brought into intimate contact.11 A new class of materials, composed entirely of “interphase”, was fabricated by forced-assembly of nanolayers

Coextrusion of a crystallizable polymer as nanolayers confines the polymer chains

to the size scale of the lamellar thickness It is anticipated that nucleation and growth habit will be affected, perhaps to the extent that new crystalline structures are created Indeed, unique row-nucleated morphologies were found in polyethylene when the layer thickness reached the dimensions of the polymer molecule.12 Unusual crystalline structures are possible in other polymers as the layer thickness is decreased from the

microscale to the nanoscale In the present study, thin isotactic polypropylene layers were coextruded between thick layers of amorphous polystyrene The solid-state structure of the crystalline polymer was studied as the polypropylene layers were made thinner and confinement by the polystyrene layers approached the nanoscale

1.2 Experimental

Microlayer and nanolayer films with 1024 alternating layers of isotactic polypropylene (PP) and polystyrene (PS) were extruded from the melt on a laboratory scale coextrusion line at Case Western Reserve University that incorporates layer-multiplying technology.13 The films were extruded onto a chill roll The take-off speed was 2 feet min-1 for 250 μm thick films and 20 feet min-1 for 40 μm thick film The total film thickness and the PP-to-PS ratio were varied to produce polypropylene microlayers and nanolayers of different thickness Films with thickness of approximately 250 μm were coextruded in compositions PP/PS of 100/0, 90/10, 80/20 and 10/90 (vol/vol) A

40 μm thick film with 10/90 composition was obtained with the faster take-off speed The polypropylene was Huntsman P4G2Z-073A with bulk density 0.900 g cm-3 and melt flow index of 1.9 g per 10 min according to ASTM D1238 The polystyrene was Dow STYRON 685D with bulk density 1.040 g cm-3 according to ASTM D 792 and melt flow index of 1.5 g per 10 min

The thickness of individual polypropylene and polystyrene layers was measured

by atomic force microscopy (AFM) The film was embedded in epoxy ChemTM/SPI-PONTM 812 KIT formulation, SPI Supplies Division of Structure Probe, Inc.) and cured for 8 hr at 60 °C Cross-sections were microtomed perpendicular to the plane of the film and observed directly The AFM images were obtained in air with a commercial scanning probe microscope (Nanoscope IIIa, Digital Instruments, Santa Barbara, CA) operated in the tapping mode Height and phase images were recorded simultaneously Measurements were performed at ambient conditions using rectangular

(SPI-type Si probes with a spring constant of 50 N m-1 and resonance frequency in the

284-362 k Hz range The tip radius was 10 nm

The morphology of individual polypropylene layers was examined by both AFM and polarizing light microscopy (OM) The film thickness was progressively reduced to several microns by peeling before the surface was examined in the polarizing optical microscope The same films were immersed in toluene for 2 hr to dissolve residual polystyrene, dried in vacuum and the surface was examined in the AFM Possible swelling of the polypropylene layer by toluene would affect only the amorphous regions, not the crystalline morphology

The crystal structure and texture of multilayered films were studied using wide angle x-ray scattering (WAXS), small angle x-ray scattering (SAXS), and the x-ray pole figure technique The WAXS system consisted of a computer controlled wide angle goniometer associated with pole figure attachment coupled to a sealed-tube source of CuKα radiation, operating at 30 kV and 30 mA The CuKα line was filtered using electronic filtering and the usual thin Ni filter The slit system that was used for collecting 2θ scans allowed for the collection of the diffracted beam with a divergence angle of less than 0.05° The following diffraction reflections from crystals of the isotactic polypropylene (PP) monoclinic modification were analyzed: the (110), (040), (130) and (060) planes containing macromolecular chains and the (-113) planes (diffraction angles 2θ of 14.1°, 16.9°, 18.5°, 25.5°, and 42.5°, respectively) The reflection from (-113) planes is the best measure of the packing and orientation of chain

axes because the monoclinic structure does not give any reflections of the (00l) type,

whereas the normal to the (-113) plane is only 5.8° away from the direction of the chain

axis, equivalent to the crystallographic c-axis The monoclinic PP crystal packing was estimated from alteration of diffraction peak positions The position of x-ray peaks was calibrated by measuring the x-ray peaks from a thin layer of graphite deposited on the surface of each sample

Measurements of the crystal orientation were performed by using a WAXS camera coupled to the x-ray generator (sealed-tube, fine point CuKα filtered source operating at 50 kV and 35 mA, Philips) using imaging plates for recording the diffraction patterns The samples were exposed in three orthogonal directions; for directions along the film a sandwich consisting of 25 pieces was prepared and glued together with isocyanate 10 s glue Exposed imaging plates were read with a PhosphorImager SI system (Molecular Dynamics)

Orientation of the crystalline phase of PP in layers was studied by means of x-ray diffraction with pole figures The procedure for pole figure data acquisition and the details of pole figure determination are described elsewhere.14 The (110), (040) and (-113) reflections from the monoclinic crystal structure of PP were analysed for the construction of pole figures The slit system of the diffractometer was selected to measure the integral intensity of the appropriate diffraction peak The necessary corrections for background scattering and sample absorption were introduced to the raw data The pole figure plots were generated by the POD program, a part of the popLA package (Los Alamos National Laboratory, Los Alamos, New Mexico) For every plot the data were normalized to the random distribution density

Lamellar orientation was probed by 2-dimensional small angle x-ray scattering (2D SAXS) The 1.1 m long Kiessig-type vacuum camera was equipped with pinhole

collimator and imaging plate as a recording medium (Kodak) The camera was coupled

to the x-ray generator (sealed-tube, fine point CuKα filtered source operating at 50kV and 35mA, Philips) The pinhole collimator allowed for resolution of scattering objects up to

50 nm Exposed imaging plates were analyzed with PhosphorImager SI system (Molecular Dynamics)

1.3 Results

Layer Characterization

Cross-sections of coextruded films with alternating layers of PP and PS are shown

in Figure 1.1 The light bands in AFM images are polystyrene layers, which had the higher elastic modulus, and the dark bands correspond to polypropylene layers, which had the lower elastic modulus Phase images of PP/PS 90/10 film in Figure 1.1a revealed continuous layers of PP and PS The texture of the crystalline PP layers somewhat obscured the layering in the low-resolution image; however, the layers were easily distinguishable at higher resolution Dark PP layers were about 460 nm thick In a

PP/PS 20/80 film (Figure 1.1b) and in a PP/PS 10/90 film (Figure 1.1c), dark PP layers

were thinner than light PS layers Good layer uniformity was clearly evident in the resolution images From higher resolution images, the PP layers were about 108 nm and

low-65 nm thick, respectively The 40 μm 10/90 film, shown in Figure 1.1d, contained PP nanolayers The PP layers, which were the thin dark lines indicated by the arrows in Figure 1.1d, were 7-13 nm thick

There was reasonably good correlation between the PP layer thickness measured from the AFM images and the estimated thickness determined from the process parameters, Table 1.1 However the thickness measured from the AFM images was consistently higher than the estimated thickness, and this trend became more pronounced

as the PP layer was made thinner Hereafter, the PP layers are identified by the layer thickness measured from AFM

The melting behavior is summarized in Table 1.2 A slight reduction in the melting temperature with decreasing layer thickness was observed A more significant

reduction in the heat of melting, corresponding to a decrease in crystallinity from 46% for the PP control to 36% for 10 nm layers, was observed as the layer thickness was reduced

Spherulitic Size Scale

The polarizing light microscope revealed 30-40 μm space-filling structures that exhibited the characteristic Maltese-cross pattern of PP spherulites, Figure 1.2 Insertion

of a ½ λ-plate produced weakly negative birefringence suggesting that the chain direction

in the α-spherulites was preferentially tangential, and correspondingly, the lamellar direction was preferentially radial In addition, a few highly negatively birefringent β-spherulites with much brighter contrast were observed

The average in-plane dimension of the spherulite-like structures as taken from

OM micrographs increased only slightly as the layers became thinner, from about 18 μm for the PP control to 34 μm, 38 μm, and 42 μm for 460 nm, 108 nm, and 65 nm films, respectively Considering that the thickness of the PP layers was orders of magnitude smaller than the in-plane dimension of the spherulite-like structures, it was reasonable to conclude that the entities observed in the OM had the shape of flattened spherulites or discoids These discoidal shapes resulted from geometric constrains imposed by the confining polystyrene layers that restricted spherulite growth to 2-dimensions Even in the 460 nm PP layers, the aspect ratio of the discoid (diameter to thickness ratio) was 70 The aspect ratio increased to 350 for 108 nm layers and to 650 for 65 nm layers The film with 10 nm PP layers exhibited only uniform and weak birefringence in the OM There was no indication of discoidal organization, Figure 1.2e

Three-dimensional AFM height images of the PP film showed the spherulite at the free surface rising to a peak in the center, Figure 1.3a This is typical of PP

crystallized as a thin film with a free surface,15 and results from volumetric changes during crystallization AFM images of exposed microlayer surfaces revealed that the PP discoids also had a peak in the center and sharp valleys where the edges impinged on

neighboring discoids, Figure 1.3b-d Because PS was above its Tg when PP crystallized,

the PS layer easily accommodated changes in interfacial shape as the PP layer crystallized The height images revealed a trend toward flatter discoids as the layer thickness decreased Films with 10 nm PP layers exhibited a smooth undulating surface with no indication of discoidal peaks and valleys, Figure 1.3e

Lamellar Size Scale

A higher resolution AFM phase image in Figure 1.4a shows the lamellar morphology of the PP control film The spherulite arms contained long radial lamellae overlaid with tangential lamellae This lamellar texture is characteristic of PP and is commonly known as cross-hatching The image in Figure 1.4b was taken from the arm

of a PP discoid in a 460 nm layer The image shows edge-on lamellae with a dense cross-hatched texture Longer radial lamellae can be distinguished from shorter

tangential lamellae However, the radial lamellae are much shorter than in Figure 1.4a

and as a result the lamellar texture appears not as well organized The lamellar morphology of discoids in 108 nm layers was virtually the same, Figure 1.4c In this case, the length of the radial lamellae, as long as 500 nm, sometimes considerably exceeded the layer thickness of 108 nm This emphasized the planarity of the discoid structure with radial lamellae preferentially oriented in the plane of the film

The morphology of 65 nm layers shown in Figure 1.4d was mostly a cross-hatched pattern consisting of thin radial and tangential lamellae The lamellae were

noticeably thinner than in the 108 nm and 480 nm layers The radial lamellae were also shorter than in 108 nm and 460 nm layers In numerous places one can distinguish lamellar fragments or globular forms of diameter of 12 – 16 nm

The lamellar texture of the 10 nm layers varied somewhat from one area to

another In some areas, exemplified by the AFM image in Figure 1.4e, fragments of

discoids were present in the form of fan-like bundles consisting of long radiating arms

2-3 μm in length The fans appeared to be randomly oriented in the plane of the nanolayer with no preferential orientation with respect to the extrusion direction Upon close inspection, the arms appeared to be closely-packed stacks of very short lamellae The fans were interspersed with areas of cross-hatched lamellae and numerous short lamellar fragments as shown in Figure 1.4f

Figure 1.5a-e shows typical SAXS patterns obtained for the control and films with

460 nm, 108 nm, 65 nm and 10 nm PP layers in the flow direction (F) Similar patterns were obtained for the other two orthogonal views: normal to the film (N) and in the transverse direction (T) The long periods determined from the position of the maximum intensity (Lorentz correction introduced) for each view and along the other two perpendicular directions in the plane of the SAXS pattern are collected in Table 1.3 In the N view, the ring of regular scattering from stacks of lamellae of rather well defined thickness were seen for the PP control, and for films with 460 nm, 108 nm, and 65 nm layers The long period was about 12.5 nm for the PP control in all directions The intensity of the peak varied when viewed along different directions, which indicated some low degree of anisotropy in lamellae orientation

Features similar to those of PP were observed for 460 nm layers when viewed in the N direction, Table 1.3 In the T and F views the long period was clearly discernible at about 12.5 nm in the F and T directions, respectively However, the T and F views showed no recognizable intensity in the N direction This last result indicated that formation of lamellae in PP layers was restricted to the edge-on position with respect to the film surface The restriction to formation of edge-on lamellae was not seen in the control PP film, only in layered films This result was confirmed by AFM images where only edge-on lamellae were seen in PP layers (Figure 1.4a-d) The value of the long period was similar to that of the PP control at about 12-12.5 nm for 108 nm layers, but decreased abruptly to 4.7-5.7 nm for 65 nm layers, Table 1.3 Evidently the spatial confinement in 65 nm layers restricted growth of edge-on 12.5 nm thick lamellae The adjustment to thinner lamellae overrode the spatial limitations

The SAXS patterns in the three orthogonal views for 10 nm nanolayers exhibited

a weakly ordered periodic lamellar structure only in the N direction, Table 1.3 The intensity in the flow direction was slightly stronger than in the transverse direction The long period was slightly larger than for 65 nm layers No long period was detected in the

T and F views for films with 10 nm layers

A strong streak was present in the center of the T and F patterns of all the layered films A streak was not present in the N direction SAXS pattern Such scattering originates from the interfaces between the PP and PS layers, which were oriented edge-on with respect to the x-ray incident beam for the T and F views only A small streak was also detected for the PP control film, which indicated incomplete fusion of the PP layers during extrusion

Crystal Structure

The 2θ scans in reflection for all multilayered films are collected in Figure 1.6 For PP and 460 nm, 108 nm, and 65 nm PP layers the peaks from monoclinic PP crystals are clearly recognizable A trace of the peak typical of hexagonal γ crystals at 2θ = 16° is visible on the shoulder of the (040) reflection of the α crystal For the 65 nm PP layer, the peaks of the monoclinic phase become less pronounced For 10 nm nanolayers only weak diffraction peaks from (040) and (110) planes are recognizable in WAXD scans

The intensity of the (040) peak for all layered films is higher than expected from the PP concentration, and it is disproportionately higher than the intensity of the peaks for other reflections This is evidence that the (040) planes are preferentially oriented in the plane of the layer, parallel to the layer interface In the case of 65 nm layers, the peaks from (110) and (040) planes are shifted towards larger values of 2θ by 0.06o and 0.09owith respect to the positions of these peaks for the PP control film The corresponding distortions of the d-spacings are from 0.6263 nm to 06235 nm for (110) planes and from 0.5240 nm to 0.5213 nm for (040) planes The peaks for 10 nm layers are shifted even more from their undisturbed positions, by 0.21o and 0.30o, respectively The d-spacing is correspondingly distorted from 0.6263 nm to 0.61736 nm for (110) planes and from 0.5240 nm to 0.51514 nm for (040) planes The changes in the d-spacings indicate that the crystals are distorted and compressed to a unit cell significantly smaller in the directions of the a- and b-axes

A change of the unit cell dimension in the direction of the c-axis can be deduced from the position and intensity of the reflection from (-113) crystal planes of the PP monoclinic crystal form The (-113) peak position in transmission does not shift but

disappears as the thickness of PP microlayers decreases, Figure 1.7 For the 65 nm and

10 nm layers the (-113) reflection cannot be resolved It means that the α crystals in these films were formed under serious spatial confinement and underwent strong distortion resulting in the loss of registry along the chains Organization of such crystals, with compressed d-spacings in the directions perpendicular to the chains and without clear registry along the chains, resembles smectic packing of conformationally distorted chains

The 2D transmission diffraction patterns of the PP control and all the layered films in the N, T, and F directions are compiled in Figure 1.8 The PP film exhibits slight orientation of (040) planes along the N and T directions The crystals are oriented with (040) planes parallel to the film surface Slight orientation of the (110) planes is also seen if viewed along the T direction This slight crystal orientation is apparently the result of the extrusion process

For the film with 460 nm layer, strong orientation of (040) planes is seen in the T and F WAXS patterns Most of the (040) planes are oriented parallel to the PP/PS interfaces In the N direction these same crystals form a strong (110) ring of uniform intensity Only a very weak (040) reflection in the N direction indicates the relative absence of crystals oriented with (040) planes in a direction other than parallel to the plane of the film Even stronger orientation of (040) planes in the film plane is indicated for 108 nm layers by equitorial concentration of (040) reflections in the T and F directions, and by absence of (040) reflections in the N direction whereas the (110) reflection is very strong

The 2D WAXS patterns in the N, T and F directions for 65 nm microlayers demonstrate a faint trace of a new crystal texture Equitorial (110) intensity in the T and

F directions indicates a crystal population with (040) planes at nearly right angles to the film plane and having chain axes parallel to the film plane

The new textural component, which appeared as a trace in 65 nm layers, is resolved in T and F direction views of the 10 nm nanolayers The (110) arcs of the new texture are sharp, occupying an azimutal angle of no more than 12-15° The (040) reflection arcs corresponding to this fraction of crystals are located at 90° with respect to the (110) arcs and they are much wider in accordance with the tilt angle between (040) and (110) planes which is 72.6° or 107.4° The sharpness of the (110) reflections indicates that this fraction of crystals originated from growth on (110) planes laying flat

well-at the PP/PS interface The amount of the new texture, estimwell-ated from the contribution to the total (110) intensity in the N view, is 10-20%

Preferred orientation of the major crystal population with (040) crystallographic planes parallel to the PP/PS interfaces is confirmed by pole figures of (110), (040) and (-113) planes for the 460 nm layers, Figure 1.9 The (-113) pole figure indicates clustering of (-113) normals parallel to the film surface, which means that the crystallized chain segments are all in the plane of the film The (110) planes are tilted in the crystal unit cell at 72.6° with respect to (040) plane, hence clustering of normals to (110) planes are expected to be around this 72.6° polar angle In fact this is seen in Figure 1.9a where the pole figure of normals to (110) planes is presented

1.4 Discussion

There are essentially two factors that influence PP crystallization in microlayers and nanolayers: the presence of PP/PS interfaces and the confined space in which crystallization takes place Any orientation effects from the extrusion process are subtle and appear only as small differences in intensity of WAXS and SAXS reflections when T and F directions are compared The effect is not strong enough to produce N direction anisotropy that is detectable by x-ray or microscopy The AFM and OM images show that crystallization is spherulitic in the PP control, and takes the form of discoids in 460,

108 and 65 nm microlayers Only fragments of discoids are present in the form of fans in

the 10 nm nanolayers (see Figure 1.4e)

The average spherulite diameter in the PP control is about 18 μm, which gives 1.71 × 105 nuclei per mm3 Given this bulk primary nucleation density of PP, the diameter of discoids in microlayers should increase significantly as the layers become thinner The 460 nm, 108 nm, and 65 nm microlayers should have discoid diameters of approximately 112 μm, 232 μm, and 300 μm, respectively However, the discoid diameter is many times smaller and is nearly the same regardless of microlayer thickness This suggests that the discoids are nucleated preferentially at PP/PS interfaces For microlayers, bulk nucleation becomes insignificant and hence the size of discoids is controlled by nucleation from interfaces In fact a weak nucleating effect of the interface

is reported for crystallization of isotactic polypropylene in a bi-layer sandwich with polystyrene.16

In surface profile, as sensed by AFM, the discoids closely resemble spherulites crystallized in thin films with a free surface with the central part of the spherulite

elevated and the peripheries lowered.15 At boundaries and at contact points between 3 spherulites, the depression is even deeper and easily visible These effects are the result

of volume contraction during crystallization.17 It is an indication that PS is above its Tg

when PP discoids form and the PS melt is able to follow the change of the interface shape

Cooling on a chill roll is rapid, and crystallization is expected to take place under conditions of high supercooling in regime III crystallization for isotactic polypropylene.18,19 Because there is no significant difference in the thermal diffusivities

of PP and PS melts, the cooling rate is nearly the same for 10 mil films regardless of composition and number of layers Faster cooling is expected for the 1 mil film Nevertheless, at high supercooling the influence of crystallization temperature on lamellar thickness should be very weak.20 Thus, systematic decreases in heat of melting and melting peak temperature of PP, 460 nm and 108 nm layers (Table 1.2) should be attributed to lower perfection of crystals and/or to higher basal energy of defective lamellar surfaces The reason for the abrupt decrease in lamellar thickness in 65 nm microlayers and 10 nm nanolayers, evident from AFM images and SAXS long period, is not clear The abrupt change to thinner lamellae is not reflected in an abrupt change in melting behavior, perhaps because the lamellar crystals have an opportunity to anneal and thicken during the relatively slow heating rate used to obtain the thermograms Edge-on orientation of lamellae in microlayers and nanolayers is inferred from AFM images and is confirmed by SAXS patterns in the T and F directions An overall decrease in strength of the long period signal as the layers become thinner indicates a loss in long distance coherence of lamellar packing

From WAXS, PP crystallizes predominantly in the α form in all microlayers and nanolayers Small traces of the β crystallographic form are visible by light microscopy The DSC heating thermograms also indicate predominantly melting of α crystals In all the layers, PP crystallizes with (040) planes parallel and (110) planes nearly perpendicular to the interface Crystallization with (040) planes lying flat is consistent with nucleation at the layer interface Studies of Lotz et al.21 indicate that (010) planes, having the lowest density of methyl groups, are most likely to be involved in epitaxial interactions

As the layer thickness decreases: (1) the (040) and (110) peaks shift towards larger values of 2θ, which is especially evident in 65 nm and 10 nm layers; (2) the (-113) peak disappears for 65 nm and 10 nm layers; (3) all other peaks disappear for 65 nm and

10 nm layers, except for the (040) and (110) peaks; and (4) a second crystal population with (110) planes parallel and (040) planes nearly perpendicular to the interface appears

in low concentration for 65 nm layers and in higher concentration for 10 nm layers

The shift in (110) and (040) peaks towards higher 2θ values indicates decreased spacing between these planes and thus tighter packing in the crystal unit cell in directions perpendicular to the chain direction Disappearance of the (-113) peak indicates worsening or even loss of registry along the chain axis It is known that tighter packing

of chains in the crystal unit cell of PP can be caused by higher isotacticity and/or higher crystallization temperature.22 However, in both cases, no shift of the (-113) peak is observed Disappearance of the (-113) peak in WAXS scans, as well as the absence of a long period in the SAXS pattern, is consistent with poor crystallographic packing in the chain direction and defective crystal surfaces Thus, tighter packing perpendicular to the

chain direction in 65 nm and 10 nm layers is accompanied by poorer order along the chain direction Even so, crystalline lamellae are distinguishable in AFM images of

65 nm and 10 nm layers along with numerous lamellar fragments or globular forms of diameter of 12 – 16 nm A similar feature found in quenched polypropylene by Keller and Norton,23 is identified with early stages of lamellae growth and branching

Thus, the thin discoids of 108 nm and 65 nm layers are largely composed of

edge-on lamellae with (040) planes lying flat edge-on the interface Isotactic polypropylene is known to crystallize as dendritic discoids that are relatively empty, i.e poorly filled with other crystals, when it is crystallized as a very thin film on mica.21 Similarly, the internal radial structure of 65 nm and 10 nm layers is difficult to fill with branching lamellae Growth of crystals with (010) planes lying flat on the interface is limited to the thickness of the layer This leaves relatively little space for secondary nucleation of the cross-hatched lamella that typically fill-in the radial structure of PP spherulites, Figure 1.10

In addition to the crystal population with (040) planes parallel to the interface, WAXS from 65 nm layers reveals another crystal fraction with (110) planes parallel to the interface and (040) planes perpendicular to the interface This fraction is more evident in WAXS of 10 nm layers, where it accounts for approximately 10-20% of the crystallinity Apparently the decrease in layer thickness results in a change of crystal growth plane from the usual (110) to the more rare (010) This is associated with significant distortion of the crystallographic unit cell including the loss of registry along the chain axis

It seems that the second population of crystals with (110) planes flat on the interface is not connected to the discoid skeleton These crystals may have nucleated at the interface or formed via homogeneous nucleation in a later stage of cooling After radial lamellae fill the 10 nm space between two PS layers via normal growth along the (110) normal direction, filling-in is realized by lamellae that nucleate with the (110) plane flat on the interface and grow in the (010) normal direction (the b-axis direction) This is possible, as indicated by Lodz and Wittman,24 and by Clark and Hoffman, although growth in the b-axis direction is slower than in the (110) normal direction Possibly the short lamellar fragments or globular structures seen in the AFM images of 65 nm layers and 10 nm layers are this fraction of crystals Nevertheless, it is not clear why PP crystals should nucleate and grow in this unusual manner when confined to nanoscale spaces.

1.5 Conclusions

This study examined the effect of thin layer confinement on crystallization of isotactic polypropylene Films with hundreds of polypropylene layers separated by amorphous polystyrene layers were fabricated by layer multiplying coextrusion The thickness of individual polypropylene layers varied from 650 nm to 10 nm Characterization of the films by conventional methods of polymer analysis revealed changes in crystal structure as the layer thickness decreased from the microscale to the nanoscale Confinement at the microscale, 0.1 μm and greater, imposed a layer thickness that was less than the spherulite dimension Under this condition, lamellae organized into flattened spherulites or “discoids” With growth in the discoid restricted to 2-dimensions, the lamellae assumed an edge-on orientation with (040) planes lying flat on the interface The normal monoclinic α crystal form was preserved Confinement at the nanoscale imposed a film thickness that was on the same size scale as the lamellar thickness In

65 nm and 10 nm thick polypropylene layers, compressed d-spacings in the directions perpendicular to the chains and loss of registry along the chain axis were suggestive of smectic packing of conformationally distorted chains In addition to the distorted crystal population with (040) planes parallel to the interface, another crystal fraction with (110) planes parallel to the interface and (040) planes perpendicular to the interface was detected

Table 1.1

Comparison of PP layer thickness measured from AFM images with PP layer

thickness calculated from processing parameters

Composition

(PP/PS vol/vol)

Film thickness (μm)

Calculated layer thickness (nm)

Measured layer thickness (nm)