MULTILAYER CO-EXTRUSION AND TWIN-SCREW COMPOUNDING OF
POLYMERIC ELASTOMER SYSTEMS
by
RONGZHI HUANG
Submitted in partial fulfillment of the requirements
for the degree of Doctor of Philosophy
Thesis Advisor: Prof. Joao M. Maia
Department of Macromolecular Science and Engineering
CASE WESTERN RESERVE UNIVERSITY
August, 2014
1
CASE WESTERN RESERVE UNIVERSITY
SCHOOL OF GRADUATE STUDIES
We hereby approve the thesis/dissertation of
Rongzhi Huang
__________
candidate for the _______Ph.D.______degree *.
(signed)________Prof. Joao Maia__________________
(chair of the committee)
_______Prof. Eric Baer__________________________
_______Prof. Liming Dai________________________
_______Prof. Kenneth Singer_____________________
(date) 06/19/2014_________
*We also certify that written approval has been obtained for any
Proprietary material contained therein.
2
DEDICATION
To my parents, Xiaolin Huang and Guihong Wang, and to my beloved wife, Yang Liu.
3
TABLE OF CONTENTS
LIST OF TABLES ...............................................................................................................7
LIST OF FIGURES .............................................................................................................8
ACKNOWLEDGEMENTS ...............................................................................................16
ABSTRACT .......................................................................................................................18
PART I: Multilayer Co-extrusion of Rheologically Mismatched Polymers
CHAPTER 1: Introduction to multilayer co-extrusion
Introduction ...................................................................................................................21
References .....................................................................................................................28
CHAPTER 2: Co-extrusion Layer Multiplication of Rheologically Mismatched Polymers:
A Novel Processing Route
Abstract .........................................................................................................................38
Introduction ...................................................................................................................39
Experimental and method .............................................................................................40
Results and discussion...................................................................................................43
Conclusions ...................................................................................................................51
References .....................................................................................................................53
Tables ............................................................................................................................55
Figures ...........................................................................................................................57
4
CHAPTER 3: Continuous Co-Extrusion of Rheologically Mismatched Polymers Using
Rectangular Multiplier Dies
Abstract .........................................................................................................................73
Introduction ...................................................................................................................73
Experimental and method .............................................................................................75
Results and discussion...................................................................................................78
Conclusions ...................................................................................................................84
References .....................................................................................................................86
Tables ............................................................................................................................89
Figures ...........................................................................................................................93
CHAPTER 4: Micro-confinement Effect on Gas Barrier and Mechanical Properties of
Multilayer Rigid/Soft Thermoplastic Polyurethane Films
Abstract .......................................................................................................................125
Introduction .................................................................................................................126
Experimental and method ...........................................................................................128
Results and discussion.................................................................................................131
Conclusions .................................................................................................................137
References ...................................................................................................................138
Tables ..........................................................................................................................141
Figures .........................................................................................................................143
PART II: Twin-screw Compounding Process for Thermoplastic Elastomer
5
CHAPTER 5: Understanding the Distribution and Dispersion of Mineral Oil in
Polypropylene/Styrene-Ethylene-Butadiene-Styrene Blends Upon Compounding
Abstract .......................................................................................................................160
Introduction .................................................................................................................160
Experimental and method ...........................................................................................162
Results and discussion.................................................................................................165
Conclusions .................................................................................................................168
References ...................................................................................................................169
Tables ..........................................................................................................................171
Figures .........................................................................................................................173
APPENDIX
Interplay Between Rheological and Structural Evolution of Benzoxazine Resins During
Polymerization
Introduction .................................................................................................................183
Experimental and method ...........................................................................................185
Results and discussion.................................................................................................187
Conclusions .................................................................................................................193
References ...................................................................................................................194
Figures .........................................................................................................................197
Bibliography ...................................................................................................................210
6
LIST OF TABLES
CHAPTER 2
Table 2.1 Relaxation spectra of PS/PMMA ......................................................................55
Table 2.2 Parameters for POLYFLOW® simulation of PS/PMMA ................................56
CHAPTER 3
Table 3.1 Geometry parameters of the square and rectangular multiplier die .................89
Table 3.2 Extrudate dimensions from different multiplier dies .......................................90
Table 3.3 Oxygen permeability of multilayer TPU films (thickness: 300 um).................91
Table 3.4 Oxygen permeability of stretched multilayer TPU films ........................................92
CHAPTER 4
Table 4.1 Comparison of integrated endothermic peak of “hard-segments domains” in
nominal 65-layer film with 75% stretching at different temperatures ............................141
Table 4.2 Mechanical properties of extruded TPU films ...............................................142
CHAPTER 5
Table 5.1 The different feeding time of PP/oil/SEBS in the internal batch mixer (unit:
second) ............................................................................................................................171
Table 5.2 Twin screw configuration: the order starts from feeder to exit die; each
conveying element is 2.4 cm, and each kneading element is 0.6 cm; every 4 kneading
elements together have the same twisting angle in arrangement ....................................172
7
LIST OF FIGURES
CHAPTER 1
Figure 1.1 Top: Layer-multiplying co-extrusion system; Bottom left: The standard
multiplier die; Bottom right: Flow stacking mechanism in the die ...................................31
Figure 1.2 The AFM phase image (left) of an EAA/PEO film with nominal PEO
individual layer thickness of 20 nm showing single lamellae crystals in PEO layer, and
the oxygen permeability of PEO layer (right)....................................................................32
Figure 1.3 The illustration of viscous encapsulation .......................................................33
Figure 1.4 The schematics of elastic instability (left) and secondary flow (right) ...........34
Figure 1.5 The interfacial instability in the multilayer film .............................................35
Figure 1.6 The elastic recoil in a square channel from the entrance (left) to exist (right) 36
Figure 1.7 The processing window of viscosity matched system (left), and mismatched
system (right) .....................................................................................................................37
CHAPTER 2
Figure 2.1 General schematics of the visualization multiplier die...................................... 57
Figure 2.2 Schematics of the cross section of the 9-layer feedblock.................................. 58
Figure 2.3 Schematics of the second-generation modular multiplier die .......................... 59
Figure 2.4 Rheological results for PS/PMMA (a), (b) and TPUs (c), (d): G’- closed
symbols, G’’- open symbols ................................................................................................... 60
8
Figure 2.5 Trouton ratio as function of Hencky strain for PS 615 (a), PMMA V100 (b),
and PMMA V826 (c); PTT fitting curves are in solid lines ................................................ 61
Figure 2.6 Trouton ratio as function of Hencky strain for Isoplast® 2530 (left) and TPU B
(right) .......................................................................................................................................... 62
Figure 2.7 Visualization results of layered structure at the end of the 2-layer feedblock
and the first generation multiplier dies (only half of the extrudate is shown due to the
symmetry) .................................................................................................................................. 63
Figure 2.8 Progression of viscosity matched (a-d) and mismatched (e-h) flow along
multiplier die for 32 layer films .............................................................................................. 64
Figure 2.9 Comparison between experiment and simulation results at the end of
multiplier die (only half of the extrudate is shown due to the symmetry) .......................... 65
Figure 2.10 N2 in the first generation multiplier die: (a) viscosity matched PS/PMMA, (b)
viscosity mismatched PS/PMMA ........................................................................................... 66
Figure 2.11 Visualization of viscosity mismatched PS/PMMA by using classical
feedblock and the first generation multiplier die (a-c; 8, 32 and 128 layers, respectively),
or 9-feedblock and the second-generation multiplier die, (d-g; 9, 17, 33 and 65 layers,
respectively) .............................................................................................................................. 67
Figure 2.12 Visualization of viscosity mismatched TPUs by using classical feedblock
and the first generation multiplier die (a-c; 5, 9 and 17 layers, respectively), or 9feedblock and the second-generation multiplier die, (d-g; 9, 17, 33 and 65 layers,
respectively) .............................................................................................................................. 68
9
Figure 2.13 Simulations of the velocity profile (top) and N2 (bottom) of viscosity
mismatched PS/PMMA ............................................................................................................ 69
Figure 2.14 Simulation results of viscosity mismatched PS/PMMA in the secondgeneration multiplier die with different wall-slip conditions (from top to bottom: 0%,
25%, 80%, 100%)...................................................................................................................... 70
Figure 2.15 Visualization results of viscosity mismatched PS/PMMA with external
lubricant: (a) feedblock, (b)-(d) from the first multiplier die (17 layers) to the third
multiplier die (65 layers) ......................................................................................................... 71
Figure 2.16 Visualization results of elasticity mismatched TPUs with external lubricant:
(a) feedblock, (b) after the first multiplier die (17 layers), (c) after the second multiplier
die (33 layers), (d) after the third multiplier die (65 layers) (e) film after the third
multiplier and coat-hanger die (65 layers). Note the significant improvement in the
layering compared to no lubricant as shown in figure 13d-g ............................................... 72
CHAPTER 3
Figure 3.1 The schematics of the square (a) and rectangular (b) multiplier die ............... 93
Figure 3.2 Rheological results for PS/PMMA (a), (b) and TPUs (c), (d): G’ is
represented by closed symbol and G’’ is represented by open symbol .............................. 94
Figure 3.3 Simulations of the velocity profile (a) and N2 (b) of PS/PMMA .................... 96
Figure 3.4 Simulation results of PS/PMMA at the output of the rectangular multiplier die
with different friction at the walls (from top to bottom: full, 1/2, 1/6, and no friction) ... 97
Figure 3.5 OM results of the extrudates of PS/PMMA from: (a) square dies, (b)
rectangular dies ......................................................................................................................... 98
10
Figure 3.6 OM results of the extrudates of TPUs from: (a) square dies, (b) rectangular
dies .............................................................................................................................................. 99
Figure 3.7 Cross-section AFM phase images of 65-layer PS/PMMA films prepared by
square (a) and rectangular (b) multiplier dies ..................................................................... 100
Figure 3.8 Analytical measurements on normalized individual layer thickness of 65-layer
PS/PMMA prepared by different multiplier dies................................................................. 101
Figure 3.9 Distribution of individual layer thickness of 65-layer PS/PMMA films
prepared by (a) square dies, (b) rectangular dies ................................................................ 102
Figure 3.10 Cross-section AFM phase images of 65-layer TPU films prepared by square
(a) and rectangular (b) multiplier dies .................................................................................. 103
Figure 3.11 Analytical measurements on normalized individual layer thickness of 65layer TPU prepared by different multiplier dies ................................................................. 104
Figure 3.12 Distribution of individual layer thickness of 65-layer TPU films prepared by
(a) square dies, (b) rectangular dies ...................................................................................... 105
Figure 3.13 OM results of the extrudates of PS/PMMA with different nominal numbers
of layers: (a) 129, (b) 257, (c) 513, (d) 1,025 ....................................................................... 107
Figure 3.14 Ratio of the lengths of flat interfaces to PS/PMMA extrudates as function of
number of layers ..................................................................................................................... 108
Figure 3.15 OM results of the extrudates of TPUs with different nominal numbers of
layers: (a) 129, (b) 257, (c) 513, (d) 1,025 ........................................................................... 110
Figure 3.16 Ratio of the lengths of flat interfaces to TPU extrudates as function of
number of layers ..................................................................................................................... 111
11
Figure 3.17 Cross-section AFM phase images of PS/PMMA multilayer films with
different nominal numbers of layers: (a) 65, (b) 129, (c) 257, (d) 513, (e) 1,025............ 114
Figure 3.18 Analytical measurements on normalized individual layer thickness of
PS/PMMA multilayer films: (a) 65, (b) 129, (c) 257, (d) 513, and (e) 1,025 layers ...... 116
Figure 3.19 Distribution of individual layer thickness of PS/PMMA films with different
numbers of layers (a) 65, (b) 129, (c) 257, (d) 513, and (e) 1,025..................................... 118
Figure 3.20 Cross-section AFM phase images of TPU multilayer films with different
nominal numbers of layers: (a) 65, (b) 129, (c) 257, (d) 513, (e) 1,025 ........................... 121
Figure 3.21 Analytical measurements on normalized individual layer thickness of TPU
multilayer films: (a) 65, (b) 129, (c) 257 layers .................................................................. 122
Figure 3.22 Distribution of individual layer thickness of TPU films with different
numbers of layers (a) 65, (b) 129, (c) 257 ............................................................................ 123
Figure 3.23 Oxygen permeability of 65 layers TPU films with different conditions. .... 124
CHAPTER 4
Figure 4.1 Rheological results of TPUs: (a) steady shear mode and (b) oscillation shear
mode; (c) extensional rheology for Isoplast® 2530, and (d) TPU B ..............................143
Figure 4.2 AFM phase images of TPU B (left), and Isoplast® 2530 (right) .................144
Figure 4.3 AFM phase images of bilayer Isoplast® 2530/TPU B: low magnification
(left), high magnification (right) .....................................................................................145
12
Figure 4.4 Morphologies of nominal 65-layer Isoplast® 2530/TPU B films: (a) OM
picture of film as extruded; AFM phase images of (b) extruded, (c) 75% stretched, and
(d) 300% stretched films .................................................................................................146
Figure 4.5 Oxygen permeability of extruded and stretched TPU films .........................147
Figure 4.6 Oxygen permeability of 75% stretched TPU films ......................................148
Figure 4.7 DSC results of extruded and stretched TPU films (heating rate=10 oC/min-1)
..........................................................................................................................................149
Figure 4.8 DSC results of 75% stretched TPU films (heating rate=10 oC/min-1) ...........150
Figure 4.9 Schematic illustration for micro-confinement effect on forming “hardsegments domain” in Isoplast® 2530 layer, and microscopic fracture at very high
deformation, in which yellow layer is Isoplast® 2530 and dark blue layer is TPU B,
orange boxes are hard segments and light blue spots are chain extenders .....................151
Figure 4.10 Normal direction 2-D WAXS patterns for various TPU films: TPU B (a),
Isoplast® 2530 (b), 65-layer (c), 65-layer with 75% stretch (d), nominal 65-layer with
300% stretch (e) ..............................................................................................................153
Figure 4.11 1-D WAXS profiles of various TPU films .................................................154
Figure 4.12 Oxygen permeability of 75% stretched nominal 65-layer film as function of
temperatures ....................................................................................................................155
Figure 4.13 DSC results of nominal 65-layer film with 75% stretching at different
temperatures ....................................................................................................................156
Figure 4.14 3-D profile on oxygen permeability of nominal 65-layer film depending on
stretching ratio and temperatures ....................................................................................157
13
Figure 4.15 Stress-strain curves for various TPU films: (a) full scale, (b) initial area
zoomed in ........................................................................................................................158
CHAPTER 5
Figure 5.1 Pictures of oil absorption in PP (a-c), and SEBS (d-f) phase at room
temperature............................................................................................................................... 173
Figure 5.2 Optical pictures of PP/oil/SEBS with different mixing time: 15-45-15 (left),
15-30-30 (right) ...................................................................................................................... 174
Figure 5.3 DMA results of PP/oil/SEBS with different mixing time .............................. 175
Figure 5.4 DMA results for extruded PP/oil (8:1) (a), SEBS/oil (8:1) (b), and SEBS/oil
(1:1) (c) blends ........................................................................................................................ 177
Figure 5.5 DMA results for the extruded TPE system after one and two compounding
cycles ........................................................................................................................................ 178
Figure 5.6 BDS of the 1st and 2nd pass samples ................................................................. 179
Figure 5.7 Oscillation shear rheology of 1st and 2nd pass samples: (a) dynamic moduli as
function of frequency; (b) complex viscosity as function of frequency ........................... 180
Figure 5.8 AFM phase images of TPE: (a) sample from 1st pass, (b) sample from 2nd pass
................................................................................................................................................... 181
APPENDIX
Figure A.1 Ring-Opening mechanism of P-ddm ............................................................... 197
Figure A.2 1H NMR spectra of P-ddm ................................................................................ 198
14
Figure A.3 Influence of polymerization temperature on polymerization kinetics for Pddm ........................................................................................................................................... 199
Figure A.4 Polymerization kinetics of P-ddm as a function of tan δ; (a) 140oC, (b) 160oC,
(c) 180oC, (d) 200oC, (e) 220oC ............................................................................................ 201
Figure A.5 Time for P-ddm to reach the final plateau in dynamic moduli ..................... 202
Figure A.6 Intermolecular hydrogen-bonding structure; (a) intermolecular H-bonding
structures between monomer and opened benzoxazine, (b) opened Mannich base H-bonds
to each other ............................................................................................................................ 203
Figure A.7 Polymerization kinetics of P-ddm as a function of temperature ................... 204
Figure A.8 Stress relaxation at different stages of the polymerization process of P-ddm
................................................................................................................................................... 205
Figure A.9 Non-isothermal DSC of P-ddm ......................................................................... 206
Figure A.10 Non-isothermal DSC of P-ddm after crosslinking ....................................... 207
Figure A.11 IR spectra of P-ddm with different polymerization stages at 140oC; (a)
Monomer, (b) 1st tan δ peak, (c) 2nd tan δ peak, (d) cured P-ddm .................................. 208
Figure A.12 IR spectra of P-ddm at different polymerization temperatures and
temperature sweep; (a) 140oC, (b) 140oC, (c) 140oC, (d) 140oC, (e) T sweep ................. 209
15
ACKNOWLEDGMENTS
First, I would like to express my sincere gratitude to my thesis advisor, Prof. Joao
Maia for his guidance, encouragement and support during my graduate studies. His
professional approach to education and research has been invaluable to my career. I
greatly appreciate to be a member of his group, learning and working in the inspirational
and cutting-edge research environment. I also want to thank other members in my thesis
committee, Prof. Eric Baer, Prof. Liming Dai, and Prof. Kenneth Singer for their time and
useful suggestions.
I would also like to appreciate the opportunity to work in the science and
technology Center for Layered Polymeric System (CLiPS) funded by National Science
Foundation, in which I have been trained professionally and socially. The generous
financial and technical support from Goodyear Tire and Rubber Company, Lubrizol
Corporation, and Saint-Gobain Company is acknowledged. Especially, I would like to
thank my collaborators, Prof. Roger Bonnecaze and Benjamin Huntington from UT
Austin, for their great supports and inputs throughout my graduate career.
Many thanks to Maia’s research group members, past and present, Jorge Silva,
Mikio Yamanoi, Alison Rodier, Arman Boromand, Chaitanya Danda, Creusa Ferreira,
Jesse Gadley, Jia Liu, Parker Lee, Patrick Harris, Ricardo Andrade, Sangjin Lee,
Seyedsafa Jamali, Shaghayegh Khani, Sidney Carson, Unique Luna, Xue Chen, for their
support, cooperation, friendship.
The Department of Macromolecular Science and Engineering has been a friendly,
interactive and social environment during my graduate studies. To Kezhen Yin, William
Lenart, Guojun Zhang, Cong Zhang, Hong Xu, Jia Wang, Shanzuo Ji, Yijian Lin, Chuan16
yar Lai, Shannon Armstrong, Zheng Zhou, Pengfei Cao, Matthew Herbert, Alex Jordan,
Joey Mangadlao, Seyedali Monemian, Sun Hua, Mingze Sun, Saide Tang, Jung-Kai
Tseng, Rocco Viggiano, Nandula Wanasekara, Amanda Way, Lianyun Yang, Tiffany
Burt, Keon-Soo Jang: thank you for contribution to our harmonious “Macro family”.
Finally, I extremely appreciate my parents and my wonderful wife for their great
support and interest in my thesis.
17
Multilayer Co-extrusion and Twin-screw Compounding of Polymeric
Elastomer Systems
by
RONGZHI HUANG
Abstract
Polymeric elastomers that possess plastic and rubber characteristics are diverse in
applications and exist in many forms, such as thermoplastic elastomer (TPE) and
thermoplastic polyurethane (TPU). According to different applications, the methods to
process these elastomers may involve extrusion, injection molding, thermoforming, and
blow molding. In this thesis, the first part mainly focuses on the co-extrusion of TPU, and
the second part is about twin-screw compounding of TPE.
Multilayer films prepared by co-extrusion have been extensively studied in the
last two decades due to the outstanding performance in gas barrier, mechanical, optical,
and dielectric properties. However, there are two limitations of the current process: first,
the co-extruded materials are limited to thermoplastics; second, the polymers to be
layered need to have the similar viscoelasticity. To broaden the processing window and
materials selection for more applications, co-extrusion of highly elastic and rheologically
mismatched polymers (e.g. TPUs) becomes necessary. In this case, there are four typical
processing instabilities needed to be solved: viscous encapsulation, elastic instability,
interfacial instability, and elastic recoil. Therefore, in the first part of this thesis, chapter 1
will briefly introduce the state-of-the-art in co-extrusion of rheologically mismatched
polymers. Specifically, chapter 2 will discuss a new processing route involving both
18
engineering and materials solutions to address the processing instabilities. However, this
approach has a limitation of producing the uniform films with more than 65 layers.
Chapter 3 will address this by incorporating the rectangular multiplier dies with aspect
ratio of 4:1. Then, the structure-properties relationship and micro-confinement effect of
the multilayer TPU films prepared by the new approach will be discussed in chapter 4.
The second part of this thesis, chapter5, will present a study on the effects of
batch and continuous mixing methods of a three component thermoplastic elastomer,
TPE, system of mineral oil polypropylene (PP) and Styrene-Ethylene-Butadiene-Styrene
(SEBS).
19
PART I: Multilayer Co-extrusion of Rheologically Mismatched Polymers
20
CHAPTER 1
Introduction to multilayer co-extrusion
1.1 Introduction
Multiphase materials with superior properties to those of one-component
polymers have been extensively studied, such as polymer blends, block co-polymers, and
nanocomposites [1-7]. Specific processing techniques involving self-assembly, in-situ
polymerization and layer-by-layer assembly are able to tailor the multiphase materials to
highly ordered structure in micro- or nano-scale, which enhance the properties and
broaden the applications [8-11]. However, these methods require solvents, are expensive,
not environmentally friendly, and are very difficult to scale up to commercial products in
industry. Melt compounding using twin-screw extruder or batch mixer is a solvent-free
process that is able to achieve co-continuous or “island-sea” morphology for polymer
blends and exfoliated fillers in nanocomposites via strong shear force. This process can
be easily scaled up, but it’s difficult to prepare the blends with ordered structure and
nanocomposites with homogeneous distribution and dispersion.
The layer-multiplying co-extrusion system that was originally developed nearly
five decades ago by Tollar James works by using either two or more extruders to feed
material into a feedblock followed by one or more multiplier dies as illustrated in Figure
1.1 [12]. This is a continuous melt-processing method that has throughput of 20 lbs./hour
and is very easily to be scaled up. In each multiplier die the layered polymers are split
vertically, compressed and expanded, and then recombined one on top of the other, which
21
doubles the number of layers [13, 14]. The process is repeated for each multiplier die. If
the feedblock produces two layers, the final number of layers after n multipliers is 2n+1.
This solvent-free process is not only cost-effective and environment-friendly, but
is also able to tune the structure of layered polymeric system from nano- to micro- scale,
which cannot be achieved by traditional polymer blends. With the current state-of-the-art,
up to 4,096 layers of thicknesses as thin as 20 nm can be co-extruded [15]. When the
ultrathin layers of crystalline polymers are prepared, such as polyethelyne oxide (PEO),
the single and high-aspect-ratio lamellae crystals are obtained due to the nanoconfinement of the other immiscible polymer layers. The single lamellae crystals are able
to change the path of the gas molecules through the film, and reduce the permeability
(oxygen) by two orders of magnitude (Figure 1.2) [15]. The mechanical properties of the
multilayer films can also be enhanced. Previous studies show that when a brittle polymer,
PEO, is layered with a ductile polymer, poly(ethylene-co-acrylic acid) (EAA), the
elongation at break of multilayer film increases to 400%, compared with 14% in single
PEO layer and 340% in single EAA layer, because the stress relief based on the
interaction of the microcraks in PEO with EAA layers is able to alleviate crazing [16, 17].
By tuning the hierarchical structure and layered configuration, the multilayer films
possess outstanding performance in gas barrier, capacitor, Bragg reflector laser, and data
storage [18-24].
However, a major limitation of the current co-extrusion process is that the
different polymer melts need to have similar viscosities and elastic properties [25]. In fact,
layer-multiplying co-extrusion is prone to a number of defects such as viscous
22
encapsulation, interfacial and elasticity instabilities when rheologically mismatched
polymers are co-extruded, which will be discussed in details in next paragraphs [26-29].
Viscous encapsulation due to the mismatched viscosities of the co-extruded
materials is the most common instability in multilayer co-extrusion. Figure 1.3 shows
that along the extrusion direction the low-viscosity layer A tends to migrate to the walls
with maximum stress and encapsulate the high-viscosity layer B [26]. The capsulation
velocity calculated from the arc length of the high-viscosity layer is the highest at the
entry of the die and decreases progressively down the channel because the low-viscosity
layer is occupying a large portion of the high shear-stress region near the walls as the
polymer melts flow down the channel [26]. Borzacchiello et al. used 3-D simulations to
further investigate the mechanism behind the viscous encapsulation in straight die [30].
Instead of isolating the viscous encapsulation from the elastic rearrangement as presented
in the previous literatures, Borzacchiello et al. conclude that the viscous encapsulation is
a second-order effect of the second normal-stress difference (N2), and there are two
regimes during the encapsulation [30]. In the first regime, once the two layers come into
the straight die and contact each other the pressure gradient pushes the low-viscosity
layer (low pressure one) until the equilibrium is achieved, the interface then shifts rapidly
because the high-viscosity layer moves as a protrusion into the low-viscosity one. This
phenomenon also depends on the layer composition and flow rate. In the second regime,
due to the unbalanced N2, the interface having the greatest absolute N2 near the walls
moves towards the midplane with the lowest N2 [30].
Secondary flow due to N2 is also a main source of the degradation in layer
structure during the multiplier process [26, 31, 32]. As Dooley points out, elastic layer
23
arrangement still occurs even that the co-extruded materials have well matched
viscosities. Figure 1.4 shows the two-layer polystyrene structure at the output of a square
channel. Since the identical materials are layering, the distorted interface is absolutely not
because of the viscous encapsulation but the result of elastic force causing the secondary
flows in the square channel. Yue et al. compared the elastic rearrangement of two
viscoelastic fluids in non-circular and circular channel by 2-D numerical simulation [32].
They found that this instability arises from two mechanisms: the first is due to the
mismatch in N2 of the two fluids, and the second is about the geometry of the channel
[32]. While a curtate cycloid interface is observed in circular channel, the less elastic
layer is wrapped by more elastic layer in non-circular die [32].
Another layer deformation in co-extrusion process is the interfacial instability
because of the interfacial slip between different layers, as shown in Figure 1.5. The
interfacial distortion can cause inhomogeneous distribution of individual layer thickness
while keeping the total thickness of the film constant [26, 33-35]. Depending on flow
rates, the interfaces might have very different morphology. For example, smooth
interfaces appear with low flow rate and wavy interfaces occur when high flow rate is
used [26, 36]. The interfacial instability is also related to extensional viscosities of the coextruded polymers. When the layers are spread in a coat-hanger die, the stretching rate
will affect the force in each layer. A wave pattern could happen when the materials with
high extensional viscosities are spreading at a high rate [26].
Last but not the least, the elastic recoil is the one that should be taken into
consideration for layer-multiplying co-extrusion. Figure 1.6 shows that even that two
24
identical polymers are co-extruded, the secondary flows at the corners of channel are
observed [26, 37].
Overcoming these processing instabilities to broaden the processing window and
materials selection range is a priority if new applications for the technology are to be
developed. In current practice, the pairs of materials must fall into one of two categories
in order for them to be usable in continuous multi-layering processes (Figure 1.7). The
first is that of materials that have an “ideal viscosity match”, which means that at some
temperature both materials must have almost equal viscosities [14]. The second category
is one where the materials must have the “off-set temperature match”, which means that
both materials must have almost matched viscosities at different temperatures but within
a 20°C range (to minimize thermal gradients and instabilities). However, the conditions
mean that up until now very high viscosity materials such as thermoplastic elastomers or
rubber, and very low viscosity materials, such as liquid crystal polymers (LCPs), could
not be co-extruded [38]. It is well known that LCPs have excellent gas barrier and optical
properties, and elastomers have ductile mechanical properties. Multilayer co-extrusion of
these two types of materials potentially might combine the advantages and enhance the
properties superior to each single one.
Micro- or nano-confinement can also benefit, if one can layer the rheologically
mismatched polymers. For example, using a high-viscosity polymer to confine the lowviscosity polymer foam is able to control the foam size, one of the most important factors
in polymer foams [39]. Nanocomposites prepared by traditional methods have difficulty
to achieve good distribution and orientation, which, on the other hand, could be improved
by layering a high-viscosity polymer to confine the filled-polymer layer [40, 41].
25
Hard/soft thermoplastics that had never been layered before could be prepared as strong
and ductile films/tubes through the new technique.
Therefore, the first part of the thesis focuses on multilayer co-extrusion of
rheologically mismatched polymers via rheology, visualization, 3-D simulation,
processing additives, re-designing the feed block, multiplier die, and coat-hanger die.
Then, once the technique is developed, the structure-properties relationship of the
multilayer rheologically mismatched polymer films will be studied.
In Chapter 2, visualization of the multilayer flows in the multiplier dies is
performed to study the mechanisms of different processing instabilities. Both engineering
and materials solutions then are proposed to expand the processing window that can be
co-extruded and layered. Viscous encapsulation is going to be minimized by optimizing
the re-designed feedblock and multiplier die to prevent crossing flows in the layers.
Secondary elastic flows are reduced by promoting wall slip via the processing aids,
essentially external lubricants. Experiments and finite element simulations (FEM)
simulations using ANSYS POLYFLOW® are used to demonstrate the success of these
approaches.
In Chapter 3, the research focuses on to further improving multilayer co-extrusion
of rheologically mismatched polymers, especially when thousands of layers are necessary
for achieving specific properties in optical lens and food packaging films [42, 43]. The
external lubricant that is immiscible with most of polymers acts as the insulating layer
between polymer melts and the walls inside the dies, and is used to significantly reduce
the second normal-stress difference (N2) at the wall and eliminate elasticity instabilities.
However, the layer thickness of external lubricant will become thinner and thinner as the
26
number of interfaces among multilayer increases. At 65 layers, the external lubricant is
too thin to exhibit the lubricating effect. Therefore, the rectangular multiplier die is
designed, which incorporates the idea of “constant cross-sectional area” and has the
aspect ratio of 4:1 at the output. A 9-layer feedblock, rectangular multiplier die system,
and external lubricant are combined to reduce processing instabilities such as viscous
encapsulation, interfacial and elastic instabilities and prepare the films with more than
one thousand layers.
In Chapter 4, a flexible soft TPU, TPU B, and an engineering TPU, Isoplast®
2530, are layered using the developed technique described in the last two chapters. TPU
B works well in applications where the polymer is required to bend and break, but it does
not have strong gas barrier properties due to its high free volume. By contrast, Isoplast®
2530 consists of 100% hard-segments and has excellent gas barrier properties, but it is
too rigid for various applications. When these two materials are multilayered, the
resulting material is aimed to possess both high gas barrier properties, and relatively good
flexibility. The extruded multilayer TPU films then are uniaxially stretched to different
amount of deformation. The structural developments in micro-scale upon stretching is
investigated and related to the gas barrier and mechanical properties.
27
1.2 References
1. Shi, W., Lynd, N. A., Montarnal, D., Luo, Y., Fredrickson, G. H., Kramer, E. J., Ntaras,
C., Avgeropoulos, A., Hexemer, A., Macromolecules, 47, 2037 (2014)
2. Liu, H., Chen, F., Liu, B., Estep, G., Zhang, J., Macromolecules, 43, 6058 (2010)
3. Moungthai, S., Mahadevapuram, N., Ruchhoeft, P., Stein, G. E., ACS Appl. Mater. &
Interfaces, 4, 4015 (2012)
4. Alig, I., Pötschke, P., Lellinger, D., Skipa, T., Pegel, S., Kasaliwal, G. R., Villmow, T.,
Polymer, 53, 4 (2012)
5. Yoo, Y., Tiwari, R. R., Yoo, Y-T., Paul, D. R., Polymer, 51, 4907 (2010)
6. Andrade, R. J., Huang, R., Herbert, M. M., Chiaretti, D., Ishida, H., Schiraldi, D. A.,
Maia, J. M., Polymer, 55, 860 (2014)
7. Litchfield, D. W., Baird, D. G., Polymer, 49, 5027 (2008)
8. Djalali, R., Samson, J., Matsui, H., J. Am. Chem. Soc., 126, 7935 (2004)
9. Charleux, B., Delaittre, G., Rieger, J., Agosto, F. D., Macromolecules, 45, 6753 (2012)
10. Kozlovskaya, V., Ok, S., Sousa, A., Libera, M., Sukhishvili, S. A., Macromolecules,
36, 8590 (2003)
11. Seo, J., Lutkenhaus, J. L., Kim, J., Hammond, P. T., Char, K., Macromolecules, 40,
4028 (2007)
12. Tollar, J. E., “Interfacial surface generator”, US patent (1966)
13. Jarus, D., Hiltner, A., Baer, E., Polym. Eng. Sci., 41, 2162 (2001)
14. Ponting, M., Hiltner, A., Baer, E., Macromol. Symp., 294, 19 (2010)
15. Wang, H., Keum, J. K., Hiltner, A., Baer, E., Freeman, B., Rozanski, A., Galeski, A.,
Science, 323, 757 (2009)
28
16. Lai, C.-Y.; Hiltner, A.; Baer, E.; Korley, L. T. J. ACS Appl. Mater. & Interfaces, 4,
2218 (2012)
17. Burt, T. M.; Keum, J.; Hiltner, A.; Baer, E.; Korley, L. T. J. ACS Appl. Mater.
Interfaces. 3, 4804 (2011)
18. Cheng, W.; Gomopoulos, N.; Fytas, G.; Gorishnyy, T.; Walish, J.; Thomas, E. L.;
Hiltner, A.; Baer, E., Nano Lett., 8, 1423 (2008)
19. Tseng, J.-K.; Tang, S.; Zhou, Z.; Mackey, M.; Carr, J. M.; Mu, R.; Flandin, L.;
Schuele, D. E.; Baer, E.; Zhu, L., Polymer, 55, 8 (2014)
20. Lai, C.-Y.; Ponting, M. T.; Baer, E., Polymer, 53, 1393 (2012)
21. Singer, K. D.; Kazmierczak, T.; Lott, J.; Song, H.; Wu, Y.; Andrews, J.; Baer, E.;
Hiltner, A.; Weder, C.; OPT. EXPRESS, 16, 10358 (2008)
22. Ryan, C.; Christenson, C. W.; Valle, B.; Saini, A.; Lott, J.; Johnson, J.; Schiraldi, D.;
Weder, C.; Baer, E.; Singer, K. D.; Shan, J. Adv. Mater., 24, 5222 (2012)
23. Lott, J.; Ryan, C.; Valle, B.; Johnson, J. R.; Schiraldi, D. A.; Shan, J.; Singer, K. D.;
Weder, C. Adv. Mater., 23, 2425 (2011)
24. Zhang, G.; Lee, P. C.; Jenkins, S.; Dooley, J.; Baer, E. Polymer, 55, 663 (2014)
25. Anderson, P. D.; Dooley, J.; Meijer, H. E. H. Appl. Rheol. 16, 198 (2006)
26. Dooley, J.: Viscoelastic Flow Effects in Multilayer Polymer Coextrusion. Ph.D thesis,
Eindhoven University of Technology, Netherlands (2002)
27. Torres, A., Hrymak, A. N., Hilton, T., Rheol. Acta, 525, 513 (1993)
28. Hatzikiriakos, S. G., Migler, K. B. Polymer Processing Instabilities: Control and
Understanding, Marcel Dekker, New York, (2005)
29
29. Nazarenko, S., Snyder, J., Ebeling, T., Schuman, T., Hiltner, A., Baer, E. SPEANTEC ‘96 Proceedings, Indianapolis, Indiana, May 5-9, (1996)
30. Borzacchiello, D., Leriche, E., Blottiere, B., Guillet, J., J. Rheol. 58, 493 (2014)
31. Debbaut, B., Avalosse, T., Dooley, J., Hughes, K., J. Non-Newtonian Fluid Mech., 69,
255 (1997)
32. Yue, P., Zhou, C., Dooley, J., Feng, J. J., J. Rheol., 52, 1027 (2008)
33. Lee, P., Dooley, J., J. Elastom. Plast. 0, 1 (2013)
34. Mahdaoui, O., Agassant, J-F., Laure, P., Valette, R., Silva, L., AIP Conference
Proceedings 907, 873 (2007)
35. Huntington, B. A., Chabert, E., Rahal, Patz, J., S., Silva, J., Harris, P. J., Maia, J.,
Bonnecaze, R. T., Int. Polym. Proc., 3, 2741 (2013)
36. Joseph, D. D., J. Non-Newtonian Fluid Mech., 70, 187 (1997)
37. Harris, P. J., Patz, J., Huntington, B. A., Bonnecaze, R. T., Meltzer, D., Maia, J.,
Polym. Eng. Sci., 54, 636 (2014)
38. Li, Z., Garza, P. A. G., Baer, E., Ellison, C. J., Polymer, 53, 3245 (2012)
39. Ranade, A. P., Hiltner, A., Baer, E., J. Cell. Plast., 40, 297, (2004)
40. Dadbin, S., Noferesti, M., Frounchi, M., Macromol. Symp., 274, 22 (2008)
41. Gupta, M., Lin, Y., Deans, T., Crosby, A., Baer, E., Hiltner, A., Schiraldi, D. A.,
Polymer, 50, 598 (2009)
42. Ji, S., Yin, K., Mackey, M., Brister, A., Ponting, M., Baer, E., Opt Eng., 52, 112105
(2013)
43. Carr, J. M., Langhe, D. S., Ponting, M., Hiltner, A., Baer, E., J. Mater. Res., 27, 1326
(2012)
30
Figure 1.1: Top: Layer-multiplying co-extrusion system; Bottom left: The standard multiplier die;
Bottom right: Flow stacking mechanism in the die [14].
31
Figure 1.2: The AFM phase image (left) of an EAA/PEO film with nominal PEO individual layer
thickness of 20 nm showing single lamellae crystals in PEO layer, and the oxygen permeability of
PEO layer (right) [15].
32
Figure 1.3: The illustration of viscous encapsulation [26].
33
Figure 1.4: The schematics of elastic instability (left) and secondary flow (right) [26].
34
Figure 1.5: The interfacial instability in the multilayer film [26].
35
Figure 1.6: The elastic recoil in a square channel from the entrance (left) to exist (right) [26].
36
Figure 1.7: The processing window of viscosity matched system (left), and mismatched system
(right) [14].
37
CHAPTER 2
Co-Extrusion Layer Multiplication of Rheologically Mismatched Polymers:
A Novel Processing Route
NOTE: Parts of this work have been submitted or published in
“Huang, R.; Patz, J.; Silva, J.; Andrade, R.; Harris, P.; Yin, K.; Huntington, B.;
Bonnecaze, R.; Cox, M.; Maia, J. M. Int. Polym. Proc., submitted”
“Maia, J., Huang, R., Cox, M., U.S. Provisional Patent Appl. No. 61/901,482 (2013)”
Abstract
In chapter 2, co-extruded films with up to 65 layers of two rheologically
mismatched polymer systems--polystyrene/poly(methylmethacrylate)(PS/PMMA) and
hard/soft thermoplastic polyurethanes (TPUs)--were successfully produced using a
combination of a 9-layer feedblock, low-pressure drop multiplier dies, and external
lubricants. Formation of viscoelastic instabilities was studied using a custom
visualization and by finite element method (FEM) simulations of a standard multiplier.
The results showed that the flow inside the standard multiplier die is highly non-uniform,
with severe gradients in shear and normal stresses and viscous encapsulation occurring
mainly in the initial multiplication stages where there is enough material available in the
low-viscosity layers to proceed with the encapsulation. To mitigate layer degradation the
standard 2- or 3-layer feedblock was replaced with a 9-layer one, thereby decreasing the
thickness of each layer at the end of the feedblock. Also, subsequent layering was
performed using a low flow resistance die. This new multiplier die yields a more uniform
flow profile and imparts a more homogeneous thermo-mechanical history on the melt
38
which results in an improved layer stability. Simulations showed that in the standard die
the second normal-stress gradients responsible for elastic instabilities at the edges of the
die are very high. These can be reduced by inducing slip at the wall resulting in be much
improved layer uniformity and stability. This was accomplished experimentally via the
use of external lubricants, and the resulting layered structure was indeed much better than
was possible to achieve with the conventional multiplier dies.
2.1 Introduction
Multilayer polymeric films have drawn much attention due to their outstanding
barrier, dielectric and optical properties [1-3]. As mentioned in Chapter 1, the major
drawback of the current process to prepare the multilayer films is that it requires the
matched viscoelasticity of the different polymer melts to be co-extruded [4-8]. There are
four types of processing instabilities, especially for rheologically mismatched polymers,
during multilayer co-extrusion: viscous encapsulation, elastic instability, interfacial
instability, and elastic recoil [9-13]. It’s necessary to understand the mechanisms behind
these instabilities and improve the process prior to new applications for the technology to
be developed.
Therefore, in this work, two solutions are proposed to expand the processing
window and the range of materials that can be co-extruded and layered. Viscous
encapsulation is minimized by optimizing the feedblock and multiplier die to prevent
crossing flows in the layers. Secondary elastic flows are reduced by promoting wall slip
via the processing aids, essentially external lubricants. Experiments and finite element
simulations (FEM) simulations using ANSYS POLYFLOW® are used to demonstrate the
success of these approaches.
39
2.2 Experimental and method
2.2.1 Materials
Two commercial poly(methyl methacrylates) (PMMA), Plexiglas® VS100 and
Plexiglas® V826 and one polystyrene (PS), Styron 615APR, were purchased from
Arkema and Styron, respectively. PMMA and PS are incompatible polymers, however,
so there is the question of whether interfacial slip plays a role in the multiplication
process, so a second system made up of chemically compatible aromatic thermoplastic
polyurethanes (TPUs), one “hard” made up almost exclusively of hard segments,
Isoplast® 2530, and one “soft” made up of 52% hard segments, TPU B, and provided by
Lubrizol Advanced Materials, Inc. were used. For some of the layering experiments, 1.5
wt% external lubricants, TR 131 or TR 251 from Struktol Company, were mixed with
PS/PMMA or TPUs [14].
These lubricants are composed primarily of unsaturated
primary amide.
2.2.2 Rheological Measurements
Steady and oscillation shear experiments were performed on a rotational
rheometer (Thermo Fisher MARS III). PS and PMMA were vacuum dried for 24 hours at
70°C, while TPUs were dried for 36 hours at 80°C prior to characterization. The
rheological experiments were conducted under a nitrogen atmosphere in order to avoid
oxidative degradation of the samples. The experiments were performed at 205oC for
TPUs or 240oC for PS and PMMA. Time-temperature superposition was performed using
the IRIS® software. The uniaxial extensional flow measurements were conducted using
40
and Sentmanat Extension Rheometer (SER) accessory [15], with sample preparation and
loading following our suggested protocol [16, 17].
2.2.3 Simulation Method
The multimode Phan-Thien-Tanner (PTT) model was applied to fit the
rheological behavior and simulate the flow of PS 615, PMMA VS 100 and PMMA V826,
which are considered to be incompressible and isothermal throughout the multiplier dies
by ANSYS POLYFLOW® [18]. Each mode of PTT model obeys the following relations:
∇
𝝉=
1
𝜆𝑖
𝛼
exp � 𝑡𝑟𝝉� + 𝜉 (𝑫. 𝝉 + 𝝉. 𝑫) = 2𝐺𝑖 𝑫
𝐺𝑖
∇ 𝜕𝝉
𝝉 = − ∇𝐯𝑇 . 𝝉 − 𝝉. ∇𝐯
𝜕𝑡
(1)
(2)
where Gi and λi are the relaxation spectra that was obtained from IRIS® calculation and
relaxation time of mode i, respectively. Even though the complete relaxation spectra of
the polymers were obtained, given the typical residence times of the melt in the multiplier
dies and in order to save the computation time, only Gi at the longest relaxation time was
used in the simulations. D is the rate of deformation tensor. The adjustable parameters (ξ
and α) control the fitting in shear and extensional flows respectively, and are kept
constant for all relaxation modes.
The viscosity η1 , is given by
η=1
(1 − ηr ) η ,
η = η 2 + η1,
(3)
(4)
41
where
ηr =
η2
.
η 2 + η1
(5)
Here the equations and parameters governing the simulation are the same as those
used in our previous works [17-19]. The parameters for the three different polymers used
in the simulations are summarized in Table 2.2.
2.2.4 Co-Extrusion System and Conditions
The system that was used to extrude the samples was comprised of two Killion
extruders, model #19782, two Zenith melt pumps, model #K46LP56, and different
numbers of multiplier dies. The velocity of the melt was controlled by the speed of the
melt pumps, which was set to 5 rpm, corresponding to melt velocities of 1x10-7 m3/s.
The PS/PMMA systems (PS 615, PMMA VS 100 and PMMA V826) were extruded at
240°C, while the TPUs (Isoplast® 2530 and TPU B) were extruded at 205°C.
A visualization multiplier die similar to the first generation multiplier die was
used to collect samples, the only difference being that the former separates into four
sections to facilitate sample removal. The appearance of the standard die flow channel is
shown in Figure 2.1. The defining characteristic of the channel is the initial asymmetrical
cross-sectional contraction of the channel, followed by a similarly asymmetrical
expansion of the area in the latter half. The samples were extracted from the visualization
die after cooling and were cut into five sections as shown in Figure 2.1. This allowed the
flow patterns to be examined at four different locations along the flow path. After cutting,
42
the cut faces of the samples were polished using 800 and then 4000 grit sandpaper. The
polished sample faces were observed under a microscope and pictures of the faces were
taken. The pictures were then used for comparison and understanding of the formation of
flow instabilities, as will be discussed in the results and discussion section.
2.2.5 9-layer feedblock and the second generation multiplier die
In order to minimize interface deformation during the early layering stages, when
the layers are thicker, a 9-layer feedblock (schematically shown in Figure 2.2) with the
same output cross section area as the standard 2-layer or 3-layer feedblock was used. This
was coupled to a new-generation modular multiplier die recently developed by our
research group, mainly by Patrick Harris, which decreases pressure drop by more than 40%
and provides a much more homogeneous thermo-mechanical history to the melt, as well
as a more symmetrical flow [17]. This new die (schematically depicted in Figure 2.3)
keeps the total cross-sectional area constant throughout a wedge-shaped channel that
imposes the simultaneous contraction and expansion of the melt; it was shown to
decrease viscous encapsulation dramatically when two rheologically mismatched
polymers are layered [17].
2.3 Results and discussion
2.3.1 Rheological results
The shear viscosity, storage and loss moduli of all materials are shown in Figure
2.4. Even though PMMA V826 shows a more pronounced shear-thinning behavior, the
shear viscosity of PMMA V826 is about one order of magnitude higher than that of PS
615 and PMMA VS100 at the low shear rates typical of multi-layering co-extrusion (~1-5
43
s-1). PMMA V826 also shows significantly higher storage moduli than both PS 615 and
PMMA VS100. The relaxation spectra of PS 615, PMMA VS 100 and PMMA V826 are
calculated by the IRIS® software and shown in Table 2.4. For the TPU systems, Isoplast®
2530 and TPU B, the viscosity ratio in the appropriate range of shear rates is over 10:1
and the elasticity ratio is higher than 100:1 in the relevant shear rate range. Due to TPU’s
complicated copolymer structure, IRIS® was not capable of calculating the relaxation
spectra, so no numerical simulations are performed for these systems.
The Trouton ratio as function of Hencky strain of the PS and PMMA melts is
depicted in Figure 2.5 PS 615 shows a relatively small strain-hardening at rates above 1
s-1, while PMMAV100 only shows strain-hardening at the highest strain rate of 10 s-1, as
does PMMA V826 but to a much smaller degree. Figure 2.6 shows the extensional
rheology for the TPU systems. While strain-hardening behavior of TPU B is observed at
each deformation rate, it shows only at the highest strain rate of 10 s-1 for Isoplast® 2530.
As shown in Figure 2.4, the PTT model fits the linear viscoelastic behavior of PS
615, PMMA VS 100 and PMMA V826 at all frequencies. For the extensional rheology, it
should be noted that the PTT model (shown in Figure 2.5 as solid lines) is able to fit the
extensional rheological results at low strain rates, but not at 10 s-1 because at this rate the
deformation is already essentially elastic and the model is no longer valid. Considering
that the strain rate and Hencky strain in the multiplier die is low, the PTT model should
be able to be applied in the numerical simulations, as per our previous works [19]. Due to
the symmetrical design of the multiplier die, only half of the multiplier die is simulated in
this work.
2.3.2 Visualization of layered structure – the standard die
44
Images were generated from all the cut and polished samples, but due to the large
resulting number of images only a select group, which shows a representative sub-group
of the data, will be discussed. In particular, samples were generated for 2, 4, 8, 16, 32, 64,
and 128 layers but for the sake of simplicity, in this paper only results for 2, 8, 32, and
128 layers will be shown.
The left column of Figure 2.7 shows the layered structure at the end of each
multiplier die for the case of viscosity matched system, and a relatively even layer
formation with increasing number of layers is observed. Important discrepancies in layer
formation to note are the uneven layer area exiting the feedblock. Conversely, the exiting
ends of the viscosity-mismatched samples can be seen on the right in Figure 2.7, and it is
clear that these are very different from the previous ones. As in the case of the viscositymatched samples, the areas of the two polymers are uneven directly after exiting the
feedblock. In the 8-layer viscosity mismatched sample, the early stages of viscous
encapsulation are already observable, with the PS phase starting to encapsulate the
PMMA phase. Another important trait of these pictures is that a void is formed in the
sample in the center, where there is a large amount of PS. This void is due to the different
cooling and contraction rates of PS and PMMA, which cause the PS phase to rupture and
distort the final solid sample (there is no indication that the voids are present in the melt).
In the 32-layer viscosity mismatched sample, the void is still present, but there is still
evidence of significant viscous encapsulation. Finally, in the 128-layer viscosity
mismatched sample, the layer thicknesses throughout the sample are still visibly uneven
and viscous encapsulation can still be observed. One interesting feature is that as the
number of layers increases it seems to stabilize the process, which was unexpected but is
45
probably due to the progressively lower amount of low viscosity material available for
encapsulation in each low-viscosity layer as they become thinner.
The evolution of layer formation and thickness was also observed along the flow
path from the beginning to the end of one multiplier for all cases, with representative
results shown in Figure 2.8 for 32-layer viscosity matched and mismatched samples. In
both cases, and since these systems begin as bilayers, there is more PS on the top of the
samples and more PMMA on the bottom. Figure 2.8 (a-d) shows that for the
rheologically matched system the layer thickness decreases in a relatively even fashion
across all the layers and some slight elastic folding occurs, especially near the center of
the die, along the entire flow path. In the flow path progression with viscositymismatched polymers, depicted in Figure 2.8 (e-h), very different results are observed.
In the beginning of the multiplier, there are 16 layers but they all have different
thicknesses, there existing large PS-only and PMMA-only areas. PS has congregated into
a thick middle layer while the PMMA accumulates in a round section on the left side of
the middle area and shows a thicker layer in the bottom half of the sample. During the
compression section of the multiplier, no real changes take place in the layer formation.
However, there are significant changes in the expansion section of the multiplier. PS
seems to migrate more towards the right wall during the expansion phase further
increasing the uneven layer formation. At the exit of the multiplier die, it can be seen that
the PMMA is not as prominent in the middle of the sample and has been replaced by PS
further increasing the unevenness of the layered structure.
Computational simulations were performed by the UT Austin partners in order to
further investigate and understand the dynamics of viscous encapsulation. As can be seen
46
in Figure 2.9, the agreement between the two is excellent, especially after the feedblock
and the first multiplier die. The first layer configuration that was investigated was the
two-layer system that is seen directly out of the feedblock. This is an important starting
point because the PS/PMMA ratios are an even 1:1 and the interface is expected to be flat.
However, both the experiments and the simulations predict an uneven distribution, with a
slight bending of the interface near the wall, which results in a larger area of PS than
PMMA. Concomitantly, the average velocities are also not the same (because the flow
rates are), with the velocity of the PMMA layer being higher. In the viscositymismatched polymers, the same phenomenon is seen in both the experimental and
simulation results.
After the viscosity-matched melts pass through one multiplier, the four resulting
layers show that the PS still occupies a slightly larger area than the PMMA, a feature that
is also shown in the simulations. In the viscosity-mismatched pair of polymers, the
interface between the two polymers has shifted and the interface is highly bent. Along the
outer walls, PS has moved up to a higher position and thus is taking up more of the cross
sectional area. This is the beginning of viscous encapsulation. In the comparison
regarding the 8-layer viscosity matched samples, the simulation is still predicting the
layer shape and thickness relatively accurately. There is not a simulation example of the
eight layer system with mismatched-viscosity polymers because of software limitations
upon the combination of the layer interfaces, which is nevertheless a sign that
encapsulation was continuing.
While looking at the progression of the layers and viscous encapsulation, the
underlying cause is not readily apparent in the samples. However, it can be observed in
47
the simulations. Dooley showed that N2 is paramount to the development of elastic
instabilities in co-extruded bi-layers in non-axisymmetric channels, and Figure 2.10
shows that N2 is also contributing significantly to layer distortion in our case [9]. In the
viscosity-matched sample, N2 remains relatively low throughout the channel, except
during the expansion section where it shows a large spike upwards. The results for the
viscosity mismatched materials show a similar trend but with a much higher value of N2
overall. It is also clear from analyzing both cases that the largest inflexion of the interface
occurs at the points where N2 is highest, which is in accordance with the previous
findings [9].
2.3.3 Engineering solution: 9-layer feedblock and second generation multiplier die
Figure 2.11 shows the layered structures resulting from the combined use of the
different feedblock and multiplier dies for the high viscosity ratio PS/PMMA system. As
can be seen, the uniformity of the layered structures through 9-layer feedblock and the
second-generation multiplier dies are dramatically improved when compared with those
obtained from the classical feedblock and the first-generation multiplier dies. Viscous
encapsulation is still very much present but progresses at a much lower rate and the
elastic instabilities also seem to be present to a much lower degree. This, in turn, has
made it possible to observe a third instability mechanism, folding of the rigid PMMA
layers near the walls. This is most likely caused by the difference in relaxation times
between the two materials. The relaxation time of PS is about one decade shorter than
that of PMMA, which also has a much higher viscosity. Thus, initially the interface bends
due to non-zero N2 and both phases deform, but upon relaxation the PMMA cannot relax
as quickly as the PS and, due to its high viscosity, it folds and causes the observed
48
morphology. Thus, while the generated layered structure is a significant improvement
upon previous state-of-the-art, is still not acceptable for practical applications. In fact,
above the limit of 33 layers, even though viscous encapsulation is controlled well by the
9-layers feedblock and the new die, interfacial and elastic instabilities are still severe and
compromise the layer structure.
As mentioned before, PS and PMMA are incompatible polymers, so low adhesion
at the interface may be increasing material mobility between layers and aiding both
viscous encapsulation and elastic instabilities. Therefore, the same experiments were
performed for a rheologically mismatched but chemically compatible pair of TPUs as
shown in Figure 2.12 (a-c). In these, both the viscosity and the elasticity ratios are higher
than those of the PS/PMMA system. Viscous encapsulation is also observed in the TPU
systems, co-extruded in a low-high-low viscosity configuration, which we will term A-BA. Figure 2.12 shows that the low viscosity TPU is already fully encapsulated after the
first multiplier die (5 layers) and the layered structure completely disappears after the
second (9 layers). As expected, improvement of the layered structure is shown when 9layer feedblock and the second-generation dies are applied. However, since the elasticity
ratio between two TPU materials is much higher than that of PS/PMMA at approximately
100:1, the elastic instabilities are even more pronounced for the TPU systems, as shown
in Figure 2.12 (d-g). In this, it is clear that the “folding” edges caused by elastic
instability begin inside the feedblock and destroy the layered structure at more than 33
layers.
Taking advantage of symmetrical geometry of the second-generation multiplier
die, 4.5 layers within half of the multiplier die were simulated for the mismatched
49
PS/PMMA systems; the results are shown in Figure 2.13. As can be seen, N2 distribute
more evenly across the die than in the standard, first-generation one (see Figure 2.10).
Additionally, the velocity profile is more consistent than that in the first multiplier die,
showing nearly flat interfaces that match reasonably well with the experiments, except
for the folding edges, which cannot be simulated since they are a consequence of the
relaxation and not of the flow.
2.3.4 Reduction of N2: the effect of external lubricant
From the results above, it is clear that if further improvement is to be
accomplished the problem stemming from non-zero N2 has to be eliminated, since it is at
the origin of the elastic instability and, indirectly, the folding of the edges. POLYFLOW®
simulations, shown in Figure 2.14, clearly show that as the degree of wall slip increases,
N2 is severely reduced and the interface between layers becomes more stable and flat.
The low die drag enables most of the elastic deformations in the melts to relax quickly,
minimizing the interfacial and elastic instabilities to a level that does not affect the layers’
structure in the second-generation multiplier die. Output profiles are zoomed in on the
bottom.
Processing aids, especially external lubricants, are able to provide extremely low
friction coefficient at the wall, and were used in this work to achieve high levels of slip at
the wall, and therefore minimize instability inception and propagation [20, 21].
Specifically, 1% of TR 251 wax and 1.5% of TR 131 wax were used for PS/PMMA or
for TPUs, respectively, during layer-multiplying co-extrusion [14]. These amounts of
external lubricant were determined experimentally as the ones that offered the best
compromise between wall-slip, which is desired, and internal lubrication, which should
50
be avoided. Figures 2.15 and 2.16 show the results with external lubricant for both the
rheologically mismatched PS/PMMA (Figure 2.15) and TPU (Figure 2.16) systems. As
can be clearly seen, using the second-generation dies in conjunction with the appropriate
amount of external results in a quite acceptable layered structure, much better than if no
external lubricant is used (Figures 2.11 and 2.12 for PS/PMMA and TPU, respectively).
In particular, the folding edges that were observed previously, are now much less
significant. Thus, using the external lubricant results in the desired (and predicted
computationally) interface stabilization effect due to the formation of an insulating layer
between the melt and the metal, which reduces N2 dramatically. This stabilization effect
is so dramatic that even for the most difficult to process system, the hard TPU/soft TPU
one, it was possible to produce final 65-layer films with a total thickness of only 360
microns (which meant the melts had to go through a further coat-hanger die after the third
multiplier). As shown in Figure 2.16 (e), the uniformity of the resulting layer structure in
the film is very acceptable. When one considers that the viscosity ratio of the TPUs is
over 10:1, the elasticity ratio is around two orders of magnitude, and the contract ratio in
coat-hanger die is extremely high, these results represent a notable success.
2.4 Conclusions
A new methodology for co-extruding multi-layered films (up to 65 layers) of
rheologically mismatched polymers is presented. Flow development was visualized to
help understand the kinetics of viscous encapsulation and elastic and interfacial
instabilities for both high viscosity and elasticity ratio polymer systems. From this
analysis it was evident that most of the low viscosity layers shift to the edge and
encapsulation of the high viscosity layers occurred during the expansion section of the
51
multiplier die. These large areas also increased the chance of voids forming in the cooled
sample. It was also observed that higher the number of layers, the slower the
encapsulation process, due to the progressively lower amount of low viscosity material
available for encapsulation. By combining a 9-layer feedblock and a new, secondgeneration multiplier die, viscous encapsulation is minimized and the uniformity of
layered structure is dramatically improved. The interfacial and elastic instabilities in
multilayer flow are further reduced by resorting to appropriate amounts of external
lubricant, which is able to form an insulating layer between the melts and metal and
reduce N2 caused by shear during processing in non-axisymmetric channels. The
simulations correctly predict the shape and position of the interface between the polymers,
and provided guidance into how much slip needs to be promoted at the wall to minimize
the phenomenon of folding of the high viscosity, high elasticity layers.
52
2.5 References
1. Wang, H., Keum, J. K., Hiltner, A., Baer, E., Freeman, B., Rozanski, A., Galeski, A.,
Science, 323, 757 (2009)
2. Mackey, M., Schuele, D. E., Zhu, L., Flandin, L., Wolak, M. A., Shirk, J. S., Hiltner,
A., Baer, E., Macromolecules, 45, 1954 (2012)
3. Lai, C.-Y., Ponting, M. T., Baer, E., Polymer, 53, 1393 (2012)
4. Carr, J. M., Mackey, M., Flandin, L., Hiltner, A., Baer, E., Polymer, 54, 1679 (2013)
5. Song, H., Singer, K., Lott, J., Wu, Y., Zhou, J., Andrews, J., Hiltner, A., Weder, C., J.
Mater. Chem., 19, 7520 (2009)
6. Jin, Y., Tai, H., Hiltner, A., Baer, E., Shirk, J. S., J. Appl. Polym. Sci., 103, 1834 (2007)
7. Ryan, C., Christenson, C. W., Valle, B., Saini, A., Lott, J., Johnson, J., Schiraldi, D.,
Weder, C., Baer, E., Singer, K., Shan, J., Adv. Mater., 24, 5222 (2012)
8. Cheng, W., Gomopoulos, N., Fytas, G., Gorishnyy, T., Walish, J., Thomas, E. L.,
Hiltner, A., Baer, E., Nano Lett., 8, 1423 (2008)
9. Dooley, J.: Viscoelastic Flow Effects in Multilayer Polymer Coextrusion. Ph.D thesis,
Eindhoven University of Technology, Netherlands (2002)
10. Anderson, P. D., Dooley, J., Meijer, H. E. H., Appl. Rheol., 16, 198 (2006)
11. Torres, A., Hrymak, A. N., Hilton, T., Rheol. Acta, 525, 513 (1993)
12. Hatzikiriakos, S. G., Migler, K. B. Polymer Processing Instabilities: Control and
Understanding, Marcel Dekker, New York, (2005)
13. Nazarenko, S., Snyder, J., Ebeling, T., Schuman, T., Hiltner, A., Baer, E. SPEANTEC ‘96 Proceedings, Indianapolis, Indiana, May 5-9, (1996)
14. Huang, R., Maia, J., Cox, M., U.S. Provisional Patent Appl. No. 61/901,482 (2013)
53
15. Sentmanat, M. L., Rheol. Acta, 43, 657 (2004)
16. Barroso, V. C., Covas, J. a., Maia, J. M., Rheol. Acta, 41, 154 (2002)
17. Harris, P. J., Patz, J., Huntington, B. A., Bonnecaze, R. T., Meltzer, D., Maia, J.,
Polym. Eng. Sci., 54, 636 (2014)
18. Thien, N. P., Tanner, R. I., J. Non-Newtonian Fluid mech., 2, 353 (1977)
19. Huntington, B. A., Chabert, E., Rahal, Patz, J., S., Silva, J., Harris, P. J., Maia, J.,
Bonnecaze, R. T., Int. Polym. Proc., 3, 2741 (2013)
20. Joseph, D. D., J. Non-Newtonian Fluid Mech., 70, 187 (1997)
21. Treffler, B., Plastics, Rubber and Composites, 34, 143 (2005)
54
Table 2.1: Relaxation spectra of PS/PMMA.
PS 615
Mode i
PMMA VS100
PMMA V826
Gi (Pa)
λI (s)
Gi (Pa)
λI (s)
Gi (Pa)
λI (s)
1
1.82 x 105
3.14 x 10-4
4.20 x 105
5.96 x 10-5
1.84 x 105
2.99 x 10-5
2
7.40 x 104
2.53 x 10-3
1.51 x 105
6.22 x 10-4
7.23 x 104
3.93 x 10-4
3
2.68 x 104
1.26 x 10-2
7.22 x 104
3.24 x 10-3
7.96 x 104
2.35 x 10-3
4
7.11 x 103
5.32 x 10-2
2.27 x 104
1.36 x 10-2
6.59 x 104
1.12 x 10-2
5
9.43 x102
2.33 x 10-1
3.08 x 103
6.02 x 10-2
3.66 x 104
5.12 x 10-2
6
3.62 x101
1.26 x 100
1.91 x 101
7.85 x 10-1
1.11 x 104
2.24 x 10-1
7
9.06 x 102
1.03 x 100
8
4.50 x 101
9.22 x 100
55
Table 2.2: Parameters for POLYFLOW® simulation of PS/PMMA
PMMA VS100
PS 615
PMMA V826
ηr (Pa.s)
0.63
0.56
0.015
η (Pa.s)
39.99
102.47
926.50
ξ
0.546
0.510
0.336
ε
0.490
0.072
0.468
λ (s)
0.785
1.260
9.220
56
Figure 2.1: General schematics of the visualization multiplier die.
57
Figure 2.2: Schematics of the cross section of the 9-layer feedblock.
58
Figure 2.3: Schematics of the second-generation modular multiplier die.
59
Figure 2.4: Rheological results for PS/PMMA (a), (b) and TPUs (c), (d): G’- closed symbols,
G’’- open symbols.
60
Figure 2.5: Trouton ratio as function of Hencky strain for PS 615 (a), PMMA V100 (b), and
PMMA V826 (c); PTT fitting curves are in solid lines.
61
Figure 2.6: Trouton ratio as function of Hencky strain for Isoplast® 2530 (left) and TPU B (right)
62
Figure 2.7: Visualization of layered structure at the end of the 2-layer feedblock and the first
generation multiplier dies (only half of the extrudate is shown due to the symmetry).
63
Figure 2.8: Progression of viscosity matched (a-d) and mismatched (e-h) flow along multiplier
die for 32 layer films.
64
Figure 2.9: Comparison between experiment and simulation results at the end of multiplier die
(only half of the extrudate is shown due to the symmetry).
65
Figure 2.10: N2 in the first generation multiplier die: (a) viscosity matched PS/PMMA, (b)
viscosity mismatched PS/PMMA.
66
Figure 2.11: Visualization of viscosity mismatched PS/PMMA by using classical feedblock and
the first generation multiplier die (a-c; 8, 32 and 128 layers, respectively), or 9-feedblock and the
second-generation multiplier die, (d-g; 9, 17, 33 and 65 layers, respectively).
67
Figure 2.12: Visualization of viscosity mismatched TPUs by using classical feedblock and the
first generation multiplier die (a-c; 5, 9 and 17 layers, respectively), or 9-feedblock and the
second-generation multiplier die, (d-g; 9, 17, 33 and 65 layers, respectively).
68
Figure 2.13: Simulations of the velocity profile (top) and N2 (bottom) of viscosity mismatched
PS/PMMA.
69
Figure 2.14: Simulation results of viscosity mismatched PS/PMMA in the second-generation
multiplier die with different wall-slip conditions (from top to bottom: 0%, 25%, 80%, 100%).
70
Figure 2.15: Visualization results of viscosity mismatched PS/PMMA with external lubricant: (a)
feedblock, (b)-(d) from the first multiplier die (17 layers) to the third multiplier die (65 layers).
71
(e)
Figure 2.16: Visualization results of elasticity mismatched TPUs with external lubricant: (a)
feedblock, (b) after the first multiplier die (17 layers), (c) after the second multiplier die (33
layers), (d) after the third multiplier die (65 layers) (e) film after the third multiplier and coathanger die (65 layers). Note the significant improvement in the layering compared to no lubricant
as shown in figure 13d-g.
72
CHAPTER 3
Continuous Co-Extrusion of Rheologically Mismatched Polymers Using
Rectangular Multiplier Dies
NOTE: Parts of this work have been submitted or published in
“Huang, R.; Chari, P.; Harris, P.; Zhang, G.; Huntington, B.; Bonnecaze, R.; Cox, M.;
Maia, J. M. Polym. Eng. Sci., submitted”
Abstract
In
this
chapter,
highly
rheologically
mismatched
poly(styrene)/
poly(methylmethacrylate) (PS/PMMA) and polyester based thermoplastic polyurethanes
(TPUs) were successfully layered to 65, 129, 257, 513, and 1,025-layer extrudates and
films using a 9-layer feedblock, rectangular multiplier dies with an aspect ratio of 4:1 at
output, and external lubricants. The extrudates directly from square and rectangular the
multiplier dies were cut and polished to study the effect of geometry on the extent of
viscous encapsulation, and elastic and interfacial instabilities via optical microscopy. The
multilayer films were characterized by atomic force microscopy (AFM) to further study
the layer uniformity and distribution of individual layer thickness after extruding from a
coat-hanger style die, and the results related to the flow evolution and processing
instabilities through the die studied by finite element method (FEM)—ANSYS
POLYFLOW®.
3.1 Introduction
Multilayer co-extrusion is not only a solvent-free process, but is also easy to be
scaled up. By tuning the hierarchical structure and layered configuration, multilayer films
73
can be made that possess outstanding performance in gas barrier, capacitor, Bragg
reflector laser, and data storage [1-7]. However, the materials to be co-extruded have to
have the matched viscoelasticity to obtain uniform layer structure, which severely
narrows the processing window and restricts applications [8, 9]. Many researchers have
performed fundamental studies, including both experiments and simulations, to
understand and overcome the viscous encapsulation and elastic instabilities during the
layer-multiplication process [10-13]. In previous chapter, it was shown that rheologically
mismatched polymers could be successfully layered to 65-layer films by re-designing the
multiplier dies and using external lubricant. By doing so, instead of squeezing and then
spreading the flow in the classical multiplier die, the cross-sectional area of a re-designed
multiplier die always keeps constant to simultaneously squeeze and spread the flow so
that the pressure drop is reduced by at least 40% [14]. The external lubricant (that is
immiscible with most polymers) acts as an insulating layer between polymer melts and
the inside walls of the dies, which promotes slip and reduces significantly the second
normal-stress difference (N2) at the wall and elastic instabilities. However, the layer
thickness of external lubricant will become thinner and thinner as the number of
interfaces among multilayer increases. At 65 layers, the external lubricant is too thin to
have a significant lubricating effect. Therefore, a solution needs to be found to further
improve this process, especially when thousands of layers are necessary for achieving
specific properties in specialized applications such as optical lens and high-performance
packaging films [15, 16].
As Dooley pointed out in his thesis, even though elastic instabilities are still
observed in a rectangular channel with the aspect ratio of 4:1, a fairly flat interface occurs
74
with distortions only near the edges of the channel, which is not the case in a square
channel [10]. Therefore, in this work we use a new constant cross-section rectangular
multiplier die with an aspect ratio of 4:1 at the outlet, combined with a 9-layer feedblock
and external lubricant, to produce films with more than 1,000 layers of highly
rheologically mismatched materials.
3.2 Experimental and method
3.2.1 Materials
Polystyrene (PS), Styron® 615APR, and poly(methyl methacrylates) (PMMA),
Plexiglas® V826, purchased from Styron company and Plexiglas company, respectively,
were used as model systems. Two aromatic and polyester-based TPUs provided by
Lubrizol Advanced Materials, Inc., ISOPLAST® 2530, consisting of 100% hardsegments and TPU B that contains 52% hard-segments were also used in this study.
PMMA and ISOPLAST® 2530 were dried at 90 oC, and TPU B was dried at 70 oC for 48
hours before co-extrusion. 0.25 wt% yellow and red color dye (PE) were added into the
resins to observe the layer uniformity of extrudates. 1.5 wt% external lubricants (TR 251,
Struktol Company) that were mainly composed by unsaturated primary amide, were dry
blended with PS/PMMA or TPUs to prepare the masterbatch for co-extrusion.
3.2.2 Rheological Measurements
All the samples were characterized via a rotational rheometry (Thermo Fisher
MARS III) according to the method introduced in Chapter 2.
3.2.3 Co-Extrusion System and Conditions
75
The multilayer co-extrusion system is described in Chapter 1, and the processing
conditions used in this study are determined based on the rheological results. Since all of
the samples are rheologically mismatched, 240 oC for PS/PMMA and 205 oC for TPUs
were chosen as the co-extrusion temperature at which viscoelasticity ratio is minimum
within the processing window. The combination of a 9-layer feedblock and the
rectangular multiplier dies (Figure 3.1) that has an output area of 2.0×0.5 inches
(5.08×1.27 cm) and an aspect ratio of 4:1 with a low pressure drop during layer
multiplication was used to produce 65, 129, 257, 513, and 1,025 layers extrudates or
films [17, 18]. In contrast, the square multiplier dies were also used to produce 65 layers
extrudates and films, and the geometrical comparison between these two types of dies is
listed in Table 3.1. The flow rate of the melt was set by the gear pumps at 10 rpm. Films
were extruded through a 10-inch coat-hanger die and collected on a stainless take-off
roller with temperature set at 80 oC. The thickness of the film is around 300 um.
3.2.4 Stretching films
Film samples with dimension of 7.6 cm × 5 cm × 0.035 cm were prepared for
uniaxial stretching. A MTS Alliance RT/30 testing machine at a gage length of 3 cm was
used to stretch the samples along the extrusion direction at 120 oC. At these temperatures,
the films were stretched to strain of 25%, 50% and 75% with deformation rate of 50% per
min. After stretching, the samples were held tightly and slowly cooled down to room
temperature.
3.2.5 Characterization
76
Extrudates directly from the rectangular multiplier dies were collected, cut and
polished using 800 and then 4000 grit sandpaper. To analyze the layer uniformity, the
polished sample faces were charaterizarized by optical microscopy (OM) involving an
Olympus (Miami, FL) BH-2 optical microscope and a CCD camera. The extruded films
from the coat-hanger die were also collected and were cut embedded and fixed in epoxy,
which is cured at room temperature for 24 hours. To prepare the sample for atomic force
microscopy (AFM), a Leica microsystmes EM FC6 ultramicrotome (Buffalo Grove, IL)
was used to cut the cross-sections of the films at -70 oC at the direction perpendicular to
extrusion. The films were examined by a Digital Laboratories Nanoscope IIIa AFM
(Digital Instruments, Santa Barbara, CA) operating in tapping mode at room temperature.
AFM phase and height images were analyzed via the NanoScope software to obtain
modulus differences and morphology information.
A MOCON OX-TRAN 2/20 (Minneapolis, MN) was used to measur the oxygen
permeability of the TPU films at 25oC, 0% relative humidity, and 1 atm pressure. Mylar
film (NIST certified) with known oxgen permeability was used to calibrate the Mocon
machine. Then, both sides of the TPU films were masked by self-adhesive aluminum
masks with a testing area of 5 cm2 at the center. Nitrogen was used to remove the
atmospheric oxygen inside the chamber for 12 hours prior to testing. The testing method
applied was referred to the work studied by Wang et al [9]. The oxygen permeability
P(O2) was calculated based on the equation below.
77
In the equation, J is the steady state flux monitored by the Mocon machine, l is the
total thickness of TPU film, D is the diffusivity, t is the flux time, and ∆p is the oxygen
pressure difference across the film (1 atm). The unit of permeability used is Barrers [9].
3.2.6 Simulation method
The Phan-Thien-Tanner (PTT) model with Gi at the longest relaxation time was
applied to fit the rheological behavior and simulate the flow of PS and PMMA, which
were considered to be incompressible and isothermal throughout the rectangular
multiplier dies by ANSYS POLYFLOW® [18]. The simulation is followed the method
performed in Chapter 2.
3.3 Results and discussion
3.3.1 Rheological properties
Rheological experiments are performed since rheology plays an important role in
multilayer co-extrusion process, which normally requires matched viscoelasticity to
obtain uniform layer structures. However, Figure 3.2 shows mismatched shear viscosities
and dynamic moduli as function of shear rate or frequency of all the materials. The shear
viscosity of PS 615 is one order of magnitude lower than that of PMMA V826 at the
range of shear rates in co-extrusion from 0.5 to 1 s-1. PS 615 also shows significantly
lower storage moduli than PMMA V826 around 0.5-1 rad/s. ISOPLAST® 2530 and TPU
B have a viscosity ratio over 10:1 and an elasticity ratio more than 100:1 in the relevant
shear rate or frequency range. It also should be noted that the PTT model fits well the
78
linear viscoelastic behavior of PS 615 and PMMA V826 at all frequencies, which enable
the simulation that will be discussed later.
3.3.2 Simulation results
POLYFLOW® simulations are able to accurately capture the evolution of the
velocity and N2 profiles of PS/PMMA flow inside the die; Figure 3.3 shows that, in
contrast to the square die, using the rectangular die results in a flat and stable interface
throughout most of the width of the extrudate, with only small distortions being observed
near the walls. This matches with the experiment results mentioned above. Our previous
work has shown that a reduction of the friction at the walls can severely reduce N2 and
improve the layer structure with stable and flat interfaces [12]. Now this wall-slip effect
is combined with the high-aspect-ratio design showed in Figure 3.4. As the friction is
lowest at the wall, interfacial and elastic instabilities are prevented and a highly uniform
layer structure is obtained.
3.3.3 Comparison between the square and rectangular multiplier die
Learning the results above, the rectangular multiplier die with aspect ratio of 4:1
at output is designed, which is combined with a 9-layer feedblock and external lubricants
to layer rheologically mismatched materials. Figure 3.5 shows the OM results of the
extrudates of PS/PMMA directly from the square (a) and rectangular multiplier (b) dies,
where the yellow layers represent PMMA the red ones are PS. While the extent of
viscous encapsulation in the extrudate of the square die is higher than that in the
rectangular die, elastic instability occurs in both. However, given the high aspect ratio of
79
the rectangular die, the distorted interfaces are left near to the wall with much more flat
and stable ones in the middle than those in the square die. Besides, as can be seen in
Figure 3.1, the input and the “squeezing and spreading” section in rectangular multiplier
dies are always kept flat and wide, while these sections in the square dies are vertical and
narrow. These factors might lead to different layer structures in the two types of dies. An
area where parts of flat interfaces are observed in Figure 3.5, and a related quantitative
analysis is shown in Table 3.2. While the length of the flat interfaces in square die is
0.42 cm accounting for 33.60% length of the whole extrudate, a much longer length
around 3.28 cm is measured in the rectangular die and accounts for 57.04% of the
rectangular extrudate. Also, it is worth noting that: a) Even though the flat surface is
relatively limited, a well-defined, if somewhat wavy layered structure is obsereved until
close to the edges; b) The exit of the multiplier dies is much thicker than the final film
and that inhomogeneities in the former will be minimized in the latter, as shall be seen
later.
This engineering solution successfully supplements the drawbacks of the material
solution—i.e. external lubricants. In fact, even though the external lubricants are able to
reduce N2 and elastic instability, the maximum loading of the external lubricants has to
be around 1.5 wt%, because beyond this percentage the pressure in the extruder will be
too low to push the melt forward in plug-flow conditions. This threshold, on the other
hand, limits the thickness of lubricating layer between the polymer melt and metal and
fails to decrease N2 enough when more than two multiplier dies are used because the
lubricating layer becomes thinner and thinner during layer multiplication (interfacial area
increases).
80
The TPU system (Figure 3.6), in which ISOPLAST® 2530 is in yellow and TPU
B is in red, shows more viscous encapsulations and elastic instability than PS/PMMA due
to the higher viscoelasticity ratio in TPUs. The length of flat interfaces and the relative
ratio are lower than those of PS/PMMA (Table 3.2) as well.
The rheologically mismatched PS/PMMA were then layered into 65-layer films
(thickness~300 µm) through the square or rectangular multiplier dies, and a 10-inch coat
hanger die. The layer structure of the cross section of the film is characterized by AFM
(Figure 3.7) where the bright area is PMMA and the dark area is PS. Unlike the
morphology observed in OM results, the interfaces between the PS/PMMA are very
straight and sharp, which is a consequence of the film flatening induced by the coathanger die. The quantitative analyses based on these AFM pictures are performed by
normalization and distribution of the individual layer thickness as shown in Figure 3.8
and 3.9. Even though the much longer flat interfaces in the rectangular multiplier die than
those in square multiplier die are observed in OM pictures, there is no big difference of
layer structures between the two types of films, in the areas of good layering. This is
attributed to that the interfaces are fully stretched horizontally in the coat-hanger die, it
does not matter how short the flat interfaces are in the multiplier die. The alternated
fat/thin layer pattern is observed in films prepared by both types of multiplier dies. For
example, the thinnest and thickest individual layer thickness from square dies are 0.86
and 9.46 µm, the average layer thickness being 4.38 µm with standard deviation of 2.42
µm. The thinnest and thickest ones from rectangular dies are 1.61 and 10.48 µm, and the
average layer thickness is 4.58 µm with standard deviation of 2.85 µm. The reasons of
such a deviation on individual layer thickness are that 1) The 9-layer feedblock has
81
different flow rates in the PMMA and PS channels, and the former flows through five
channels, while the latter flows through four; 2) The low-viscousity PS tends to flow
towards the edge encapsulating the high-viscosity PMMA, despite the use of a balanced
multiplier die and external lubricants, a problem that is especially relevant in the coathanger die, which has a large contract ratio.
Figure 3.10 shows the cross-section AFM phase images of the 65-layer TPU
films (thickness ~400 µm). A well-defined multilayer structure is obtained even though
the viscoelasticity ratio of two TPUs is around 2 orders of magnitude. The alternated
fat/thin layer pattern is also observed in TPU films, and is more pronounced than that of
PS/PMMA, because the former has higher viscoelasticity ratio than the latter. It should be
noted that for TPUs the distribution of individual layer thickness from rectangular dies is
more homogeneous than that from square dies. As shown in Figure 3.11 and 3.12, even
though the individual layer thicknesses of both square and rectangular dies concentrate at
the range of 4~6 µm, the results of square dies show more thick layers of 10~20 µm.
While the average layer thickness from the square die is 6.20 µm with a standard
deviation of 5.03 µm, the average layer thickness from the rectangular die is slightly
more unform at 6.41 µm with a standard deviation of 4.56 µm. This is because that for a
very high viscoelasticity ratio system, the flat input and simultaneous “squeezing and
spreading” sections in rectangular dies are able to maintain a more homogeneous layer
structure than square dies during the layer-muliplication process.
3.3.4 Extrudates with different numbers of layers in rectangular multiplier die
82
With the advantage of the rectangular multiplier die established in terms of
relative workable area relatively to total film width, and of layer uniformty for the TPU
systems, extrudates with higher numbers of layers were prepared using this multiplier. As
can be seen in Figure 3.13, the viscous encapsulation, interfaces and elastic instabilities
are progressively reduced and the layered structure becomes more and more uniform as
the numbers of the layers increases, due to the increased confinement, which leaves less
material available in each layer to propagate the instabilities. For example, at 1,025 layers,
the individual layer thickness is around 12 µm, befeore spreading in the coat-hanger die.
It should be noted that starting from 513 layers, it seems like there is no big difference
between viscosity matched and mismatched system, which is in agreement with our prior
findings (chapter 2). This can be better explained by analizing Figure 3.14, which shows
that the ratio of the length of the flat interfaces to the extrudate (a/b) increases from 57.04%
to 87.50% cm as the number of layers increases from 65 to 1,025.
Figure 3.15 shows the OM results for TPU extrudates with numbers of layers
from 129 to 1,025. The layer uniformity clearly improves visually as the number of layers
increases, showing the same tendency observed in PS/PMMA. It should be noted that
some of the interfaces in the 1,025-layer sample are blurred with some layer break-up.
This is attributed to the interdiffusion between the two chemical compatible TPUs when
the individual layer thickness achieves micro-scale, which will be discussed in details
below. Figure 3.16 also shows that the ratio a/b in TPUs are lower than those of
PS/PMMA at all numbers of layers. For both PS/PMMA and TPUs, 513-layer sample
correponding to individual layer thickness of 25 µm is the threshold value to obtain
83
highly uniform multilayer structure when rheologically mismatched polymers are coextruded.
3.3.5 Films with different numbers of layers using the rectangular multiplier die
Once the layered structure of the extrudates was understood, thin multilayered
PS/PMMA and TPU films with increasing numbers of layers were also prepared via the
rectangular multiplier dies. AFM phase images (Figure 3.17) show that continuous layers
with straight and sharp interfaces are obtained in all PS/PMMA samples. Although the
alternated fat/thin layer pattern is still observed, the layer structure becomes more and
more uniform as the number of the layers increases showing the same trend as observed
in extrudates. Figure 3.18 and 3.19 show that while at 65 layers the distribution of the
individual layer thickness is very broad and discrete, a Gaussian distribution of the
individual layer thickness concentrated at 0.3 µm is formed at 1,025 layers. This indicates
a uniform multilayer structure is formed because the measured layer thickness of 300 nm
is very close to the 293 nm nominal thickness of the 1,025-layer film.
The multilayer structures of TPU films from 65 to 1,025 layers are shown in
Figure 3.20. Due to the higher viscoelasticity ratio and interdiffusion, TPU films have
worse layer uniformity than PS/PMMA, and there is no well-defined layer structure
observed in the 513 and 1,025 layers film. Figure 3.21 and 3.22 show that the 65- and
129-layer TPU films have discrete distribution of inividual layer thickness and 257-layer
has periodical ultra-fat/thin layer structure. According to our previous study focused on
interdiffusion of co-extruded TPU, the interphase between two aromatic and polyester
based TPU has a thickness around 2 µm [19]. Thus, starting from a 129-layer TPU film
corresponding to nominally individual layer thickness of 2.3 µm, the interdiffusion
84
between two TPUs becomes very pronounced and they diffuse into each other when 513
and 1,025-layer film are produced.
3.3.6 Gas barrier properties of TPU films
Gas (oxygen) permeabilities of TPU films are shown in Table 3.3. The multilayer
films show decreased permeability as the number of the layers increases, even though the
reduction is very small. The TPU chains, especially in glassy and amorphous Isoplast®
2530, are oriented during the squeezing and spreading process in slit multiplier dies,
which leads to forming the impermeable “hard-segments domains” through
intermolecular hydrogen bondings along the backbones that contain “–NH” and “C=O”
groups. On the other hand, the interdiffusions partially ruin the impermeable structures
when the individual layer thickness is reduced to a few microns. Then, the microconfinement effect on gas barrier properties is affected, and the gas permeability
decreases by a very small amount only.
Uniaxial stretching at 120 oC further helps orienting the TPU chains and forming
the impermeable “hard-segments domains”. Table 3.4 shows a decreased permeability as
the percentage of stretching increases for 65 layers. However, the reversed trend is
observed for 129 and 257 layers, which, again, is due to the interdiffusion of TPUs.
In order to exam the enhanced barrier properties that are due to uniaxial stretching,
a controlled test, annealing the 65 layers TPU film at 120 oC (with the same time of
stretching), is performed. In Figure 3.23, both the as-extruded and the annealed 65 layers
films show the same permeability of 0.067 barrer, but the 75% stretched sample shows a
50% improvement of 0.031 barrer.
85
3.4 Conclusions
In this study, PS/PMMA with viscoelasticity ratio of 10 and TPUs with
viscoelasticity ratio of 100 were successfully co-extruded to 65, 129, 257, 513, and
1,025-layer extrudates and films through a 9-layer feedblock, the rectangular multiplier
dies with an aspect ratio of 4:1 at output, and external lubricants. OM results show that
the extrudates from rectangular multiplier die have longer, more stable and flatter
interfaces than the ones from square die. The layer uniformity and distribution of
individual layer thickness of multilayer films improve as the number of the layers
increase due to the less amount of low viscous material at the edge available to
encapsulate the high viscous one and micro-confinement effect. Simulation results show
that the combination of high-aspect-ratio output and low friction at the wall help form the
uniform layer structure. Due to inter-diffusion between the two polyester based TPUs,
257-layer film is the top limit to obtain a multilayer structure in TPU and 129 layers is
the limit for which good, uniform layering is observed.
86
3.5 References
1. Cheng, W.; Gomopoulos, N.; Fytas, G.; Gorishnyy, T.; Walish, J.; Thomas, E. L.;
Hiltner, A.; Baer, E., Nano Lett., 8, 1423 (2008)
2. Tseng, J.-K.; Tang, S.; Zhou, Z.; Mackey, M.; Carr, J. M.; Mu, R.; Flandin, L.; Schuele,
D. E.; Baer, E.; Zhu, L., Polymer, 55, 8 (2014)
3. Lai, C.-Y.; Ponting, M. T.; Baer, E., Polymer, 53, 1393 (2012)
4. Singer, K. D.; Kazmierczak, T.; Lott, J.; Song, H.; Wu, Y.; Andrews, J.; Baer, E.;
Hiltner, A.; Weder, C.; OPT. EXPRESS, 16, 10358 (2008)
5. Ryan, C.; Christenson, C. W.; Valle, B.; Saini, A.; Lott, J.; Johnson, J.; Schiraldi, D.;
Weder, C.; Baer, E.; Singer, K. D.; Shan, J. Adv. Mater., 24, 5222 (2012)
6. Lott, J.; Ryan, C.; Valle, B.; Johnson, J. R.; Schiraldi, D. A.; Shan, J.; Singer, K. D.;
Weder, C. Adv. Mater., 23, 2425 (2011)
7. Zhang, G.; Lee, P. C.; Jenkins, S.; Dooley, J.; Baer, E. Polymer, 55, 663 (2014)
8. Ponting, M., Hiltner, A., Baer, E., Macromol. Symp., 294, 19 (2010)
9. Wang, H., Keum, J. K., Hiltner, A., Baer, E., Freeman, B., Rozanski, A., Galeski, A.,
Science, 323, 757 (2009)
10. Dooley, J.: Viscoelastic Flow Effects in Multilayer Polymer Coextrusion. Ph.D thesis,
Eindhoven University of Technology, Netherlands (2002)
11. Yue, P., Zhou, C., Dooley, J., Feng, J. J., J. Rheol., 52, 1027 (2008)
12. Huntington, B. A., Chabert, E., Rahal, Patz, J., S., Silva, J., Harris, P. J., Maia, J.,
Bonnecaze, R. T., Int. Polym. Proc., 3, 2741 (2013)
13. Anderson, P. D., Dooley, J., Meijer, H. E. H., Appl. Rheol., 16, 198 (2006)
87
14. Harris, P. J., Patz, J., Huntington, B. A., Bonnecaze, R. T., Meltzer, D., Maia, J.,
Polym. Eng. Sci., 54, 636 (2014)
15. Ji, S., Yin, K., Mackey, M., Brister, A., Ponting, M., Baer, E., Opt Eng., 52, 112105
(2013).
16. Carr, J. M., Langhe, D. S., Ponting, M., Hiltner, A., Baer, E., J. Mater. Res., 27, 1326
(2012).
17. Barroso, V. C., Covas, J. a., Maia, J. M., Rheol. Acta, 41, 154 (2002)
18. Thien, N. P., Tanner, R. I., J. Non-Newtonian Fluid mech., 2, 353 (1977)
19. Silva, J., Maia, J. M., Huang, R., Meltzer, D., Cox, M., Andrade, R., Rheol Acta, 51,
947 (2012).
88
Table 3.1: Geometry parameters of the square and rectangular multiplier die.
Multiplier Die
Inlet (inch)
Outlet (inch)
Aspect ratio
Square die
¼×½
½×¼
1:1
Rectangular die
1 ×½
2×¼
4:1
89
Table 3.2: Extrudate dimensions from different multiplier dies
Length of flat
Length of extrudates
Ratio of a and b
interfaces (a)
(b)
(%)
PS/PMMA 65L square
0.42
1.25
33.60
PS/PMMA 65L rectangular
3.28
5.75
57.04
TPUs 65L square
0.43
1.25
34.40
TPUs 65L rectangular
3.05
5.15
59.22
90
Table 3.3: Oxygen permeability of multilayer TPU films (thickness: 300 um)
Number of layers
Norminal individual layer
Permeability
thickness (um)
(Barrer)
65
4.61
0.068
129
2.32
0.067
257
1.16
0.065
513
0.58
0.055
1,025
0.29
0.053
91
Table 3.4: Oxygen permeability of stretched multilayer TPU films
Multilayer films with different
Permeability (Barrer)
percentage of stretching
65L 25%
0.051
65L 50%
0.046
65L 75%
0.031
129L 25%
0.055
129L 50%
0.060
129L 75%
0.062
257L 25%
0.058
257L 50%
0.064
257L 75%
0.070
92
(a)
(b)
Figure 3.1: The schematics of the square (a) and rectangular (b) multiplier die.
93
Figure 3.2: Rheological results for PS/PMMA (a), (b) and TPUs (c), (d): G’ is represented by
closed symbol and G’’ is represented by open symbol.
94
(a)
95
(b)
Figure 3.3: Simulations of the velocity profile (a) and N2 (b) of PS/PMMA
96
Figure 3.4: Simulation results of PS/PMMA at the output of the rectangular multiplier die with
different friction at the walls (from top to bottom: full, 1/2, 1/6, and no friction).
97
(a)
(b)
Figure 3.5: OM results of the extrudates of PS/PMMA from: (a) square dies, (b) rectangular dies.
98
(a)
(b)
Figure 3.6: OM results of the extrudates of TPUs from: (a) square dies, (b) rectangular dies.
99
(a)
(b)
Figure 3.7: Cross-section AFM phase images of 65-layer PS/PMMA films prepared by square (a)
and rectangular (b) multiplier dies.
100
Figure 3.8: Analytical measurements on normalized individual layer thickness of 65-layer
PS/PMMA prepared by different multiplier dies.
101
(a)
(b)
Figure 3.9: Distribution of individual layer thickness of 65-layer PS/PMMA films prepared by (a)
square dies, (b) rectangular dies.
102
(a)
(b)
Figure 3.10: Cross-section AFM phase images of 65-layer TPU films prepared by square (a) and
rectangular (b) multiplier dies.
103
Figure 3.11: Analytical measurements on normalized individual layer thickness of 65-layer TPU
prepared by different multiplier dies.
104
(a)
(b)
Figure 3.12: Distribution of individual layer thickness of 65-layer TPU films prepared by (a)
square dies, (b) rectangular dies.
105
(a)
(b)
(c)
106
(d)
Figure 3.13: OM results of the extrudates of PS/PMMA with different nominal numbers of layers:
(a) 129, (b) 257, (c) 513, (d) 1,025.
107
Figure 3.14: Ratio of the lengths of flat interfaces to PS/PMMA extrudates as function of number
of layers.
108
(a)
(b)
(c)
109
(d)
Figure 3.15: OM results of the extrudates of TPUs with different nominal numbers of layers: (a)
129, (b) 257, (c) 513, (d) 1,025
110
Figure 3.16: Ratio of the lengths of flat interfaces to TPU extrudates as function of number of
layers.
111
(a)
(b)
112
(c)
(d)
113
(e)
Figure 3.17: Cross-section AFM phase images of PS/PMMA multilayer films with different
nominal numbers of layers: (a) 65, (b) 129, (c) 257, (d) 513, (e) 1,025.
114
(a)
(b)
(c)
115
(d)
(e)
Figure 3.18: Analytical measurements on normalized individual layer thickness of PS/PMMA
multilayer films: (a) 65, (b) 129, (c) 257, (d) 513, and (e) 1,025 layers.
116
(a)
(b)
(c)
117
(d)
(e)
Figure 3.19: Distribution of individual layer thickness of PS/PMMA films with different
numbers of layers (a) 65, (b) 129, (c) 257, (d) 513, and (e) 1,025.
118
(a)
(b)
119
(c)
(d)
120
(e)
Figure 3.20: Cross-section AFM phase images of TPU multilayer films with different nominal
numbers of layers: (a) 65, (b) 129, (c) 257, (d) 513, (e) 1,025.
121
(a)
(b)
(c)
Figure 3.21: Analytical measurements on normalized individual layer thickness of TPU
multilayer films: (a) 65, (b) 129, (c) 257 layers.
122
(a)
(b)
(c)
Figure 3.22: Distribution of individual layer thickness of TPU films with different numbers of
layers (a) 65, (b) 129, (c) 257.
123
Figure 3.23: Oxygen permeability of 65 layers TPU films with different conditions.
124
CHAPTER 4
Micro-confinement Effect on Gas Barrier and Mechanical Properties of Multilayer
Rigid/Soft Thermoplastic Polyurethane Films
NOTE: Parts of this work have been submitted or published in
“Huang, R.; Chari, P.; Tseng, J-K.; Zhang, G.; Cox, M.; Maia, J. M. J. Appl. Polym. Sci.,
submitted”
Abstract
Rigid/soft thermoplastic polyurethane (TPU) films were produced via layermultiplying co-extrusion and the effect of confinement on morphology and gas barrier
and mechanical properties is studied. The soft TPU, which has 52% hard-segments,
shows phase separation, while the rigid TPU with 100% hard-segments exhibits
amorphous structures. Then, the multilayer TPU films are uniaxially stretched to different
amounts of deformations, from 0% to 300%. Even though the viscosity ratio of the two
TPUs is over 10 and the elasticity ratio around 100, optical and atomic force
microscopies show that a multilayer structure is successfully achieved. DSC and WAXS
results show that micro-confinement occurs during orientation, upon which a significant
reduction in oxygen permeability of multilayer TPU film is observed when the films are
stretched at 75% when compared to the mono and bi-layer TPU. In the meanwhile, the
dependence of gas barrier properties on temperature and deformation is also investigated.
Stress-strain curves of TPU films are obtained through MTS tensile machine, and 100%
improvement on elongation at break is found when compared monolayer to multilayer
TPU film.
125
4.1 Introduction
Thermoplastic polyurethane (TPU) is an important elastomer exhibiting high melt
strength, good mechanical properties, and excellent abrasion resistance, with many
intricate applications ranging from biomaterials to footwear [1-4]. By tuning the
composition of hard-segments and soft-segments during TPU synthesis, the material
created can show a wide range of properties. It is well known that the gas barrier and
mechanical properties of TPU depend on the amounts of hard-segments, which have a
higher glass transition temperature than soft-segments [5-7]. The issue with the particular
material is that with more hard-segments, the material becomes more rigid, and as a result,
is not flexible enough for a number of applications, e.g., packaging. Soft-segments play
an important role in phase separation via different chain lengths, chain extender
structures and molecular weight of soft segments [8, 9]. When analyzing the balance of
mixing relative fractions of hard and soft phases and its effect on the glass transition
temperature, studies have shown that hydrogen bonding with ester groups is stronger than
with ether groups [10]. Thus, by integrating ester groups, a stronger and more flexible
material can be produced. Finding a good balance between these two phases will yield a
material that has hard and soft phases, but also flexibility needed for various applications
[7-10]. For example, if a TPU can be designed to contain both good gas barrier and
flexibility characteristics, then devices such as biomedical bags/tubes can be produced.
Creating multilayered polymeric systems through co-extrusion and biaxial
stretching yields favorable characteristics, such as ultra-low gas permeability, high
dielectric performance, gradient refractive index lens, 1-D photo crystals, and 3-D storage
[11-20]. The purpose of using this layer-multiplying co-extrusion is to control the layer
compositions and thickness of two polymers, and in doing so, allowing the maximization
126
of the effect of hierarchical structures and nano-confinement of the polymers [21-25].
Such an effect can also be achieved by adding nano-fillers into polymer matrix forming
nanocomposites. For example, Dekun et al. have shown that adding organically modified
nanoclays into TPU can decrease the gas permeability of the nanocomposites [26].
However, several issues have to be pointed out: first, such an improvement on barrier
properties requires a very high amount of nanoclays, e.g. 10 wt%, which, on the other
hand, decreases the elongation at break and tensile strength of TPU. Second,
homogeneous distribution and exfoliated dispersion of nanoclays in TPU system is very
difficult to achieve, especially at high contents of nanoclays where lots of agglomerations
appear. The nano-confinement of polymer chains due to multi-layered co-extrusion is
able to not only improve the strength of the material, but also significantly enhance the
gas barrier properties especially when the individual layer thickness is around 30 nm, and
single crystalline orientation is created [27-29]. This indicates that with a multilayer
structure, the orientation of the crystalline structure is altered and can be manipulated to
improve the mechanical properties and barrier properties of the TPU while maintaining
the flexibility.
Even though the structural changes at the micro-scale upon uniaxial stretching of
TPU with various ratios of hard-segments are extensively studied, the uniaxial stretching
on multilayer TPU under micro-confinement has never been studied [30, 31]. In this
paper, a flexible soft TPU, TPU B, and an engineering TPU, Isoplast® 2530, are
multilayered by co-extrusion. TPU B works well in applications where the polymer is
required to bend and break, but it does not have strong gas barrier properties due to its
high free volume. By contrast, Isoplast® 2530 consists of 100% hard-segments and has
127
excellent gas barrier properties, but it is too rigid for various applications. When these
two materials are multilayered, the resulting material is aimed to possess both high gas
barrier properties, and relatively good flexibility.
4.2 Experimental and method
4.2.1 Materials
Two aromatic and polyester-based TPUs provided by Lubrizol Advanced
Materials, Inc. were used in this study. A rigid TPU, Isoplast® 2530 consisted of 100%
hard-segments, and a soft TPU B with 52% hard-segments were dried at 80 oC for 24
hours prior to processing. 1.5 wt% external lubricants (TR 251) provided by Struktol
Company, which mainly compose unsaturated primary amide, were added into TPU
during co-extrusion [32, 33]. Extensional rheometry experiments were performed in a
SER device coupled to a Paar Physica MCR 501 rheometer. Sample preparation and
loading followed the procedure recommended by Barroso et al., in order to ensure the
samples were stress-free in not sagging at the beginning of the experiments [34].
4.2.2 Rheological measurement
The co-extrusion processing conditions were determined based on the shear
rheological properties of Isoplast® 2530 and TPU B charaterized by rotational rheometry
(Thermo Fisher MARS III). Since these two TPUs are rheologically mismatched, 205 oC
was chosen as the co-extrusion temperature at which viscoelasticity ratio is minimum
within the processing window. The combination of a 9-layer feedblock and a second
generation multiplier dies recently developed by the authors is used to produce 65-layer
TPU films [32, 33, 35]. The flow rates of the TPU melts were the same and set by
128
identical gear pumps. Films were extruded through a 3 inches coat-hanger die and
collected on a stainless take-off roller with temperature set at 80 oC. Mono-layer and bilayer Isoplast® 2530 and TPU B films of 175 microns were also produced.
4.2.3 Stretching films
Film samples with dimension of 7.6 cm × 5 cm × 0.035 cm were prepared for
uniaxial stretching. A MTS Alliance RT/30 testing machine at a gage length of 3 cm was
used to stretch the samples along the extrusion direction at 100 oC, 110 oC, 120 oC, and
130 oC. At these temperatures, the films were stretched to strain of 25%, 50%, 75%,
100%, 200%, and 300% with deformation rate of 50% per min. After stretching, the
samples were held tightly and slowly cooled down to room temperature.
4.2.4 Characterization
The mechanical tests of TPU films, stress-strain curves, were performed with a
MTS Alliance RT/30 testing machine. TPU films were cut into “dog-bone” shape with
dimension of 0.44 cm × 0.035 cm and uniaxially stretched until break at room
temperature with deformation rate of 50% per min.
Differential scanning calorimetric (DSC, TA Instruments Q-100) measurements
were performed on both as-extruded and stretched TPU films ranging from -50 to 250 oC
at a heating/cooling rate of 10 oC per min with sample sizes of 5-8 mg.
Two dimensional wide angle X-ray scattering (WAXS) measurements were
carried out on mono-, bi-, nominal 65-layer films with different percentage of stretching
in the normal direction (ND) to characterize the molecular structure and orientation of the
TPUs. WAXS experiments were performed under vacuum and at room temperature (25
129
o
C), with an X-ray beam based on highly focused monochromatic CuKa (λ= 0.1542 nm)
generated from a micro-focus X-ray generator (Rigaku, MicroMax-002, Woodlands, TX)
that was equipped with two laterally graded multilayer optics side-by-side. This
collimated monochromatic X-ray beam was operated at 45 kV and 0.88 mA by using
three pinholes, with the diameter of the beam around 700 mm. The images then were
taken by using Fujifilm magnetic imaging plates and were processed by a Fujifilm FLA7000 image plate reader. In order to obtain strong reflection patterns, extruded and
stretched films were exposed for 8 and 16 h, respectively.
TPU films were embedded and fixed in epoxy that is cured at room temperature
for 1 day. A Leica microsystmes EM FC6 ultramicrotome (Buffalo Grove, IL) was used
to microtome the cross sections of TPU films at -70 oC with direction perpendicular to
extrusion. The layer uniformity of extruded and strectched 65-layer TPU films were
charaterized by optical microscopy (OM) with an Olympus (Miami, FL) BH-2 optical
microscope and a CCD camera. Mono-layer TPU films were examined with a Digital
Laboratories Nanoscope IIIa AFM (Digital Instruments, Santa Barbara, CA) operating in
tapping mode at room temperature. AFM phase and height images were analyzed via the
NanoScope software to obtain modulus differences and morphology information.
The oxygen permeability measurements of TPU films were conducted by a
MOCON OX-TRAN 2/20 (Minneapolis, MN) at 25 oC, 0% relative humidity, and 1 atm
pressure. Prior to testing, Mylar film (NIST certified) with known oxgen permeability
was used to calibrate the Mocon machine. Then, both sides of the TPU films were
masked by self-adhesive aluminum masks with a testing area of 5 cm2 at the center.
Nitrogen was used to remove the atmospheric oxygen inside the chamber for 12 hours.
130
The oxygen permeability P(O2) was calculated from the equation below:
P(O2) = J
𝑙
∆P
in which, J is the steady state flux minitored by Mocon machine, l is the total thickness of
TPU film, and ∆P is the oxygen pressure difference across the film (1 atm). The unit of
permeability used in this study is Barrers.
4.3 Results and discussion
4.3.1 Rheological properties
The rheologically mismatched hard/soft TPUs with viscosity ratio over 10 and
elasticity ratio of 100 at shear rate/frequency of 1 s-1 during co-extrusion [Figure 4.1 (ab)], were successfully layered to 65-layer films with dimension of 7.5 cm (width) × 0.035
cm (thickness). Extensional rheological tests are also performed on these two TPUs. In
Figure 4.1 (c-d), TPU B shows strain-hardening behavior at all deformation rates, while
Isoplast® 2530 mostly exhibites strain-softening except at the highest deformation rate of
10 s-1. The reason for the high melt strength of TPU B is the long-range order caused by
phase seperation between soft and hard segments. The strong intermolecular interaction
leads to high extensional viscosity during stretching. On the other hand, Isoplast® 2530,
as an amorphous glassy TPU, consists of bulky polymer chains and yields to the
deformation quickly, except at a very high deformation rate, where long-range,
crystalline-like structures develop [36].
4.3.2 Morphology
131
The morphologies of mono-, bi-, and multilayer TPUs films are charaterized by
OM and AFM. As shown in Figure 4.2, while TPU B shows the long-range orders in
which bright areas represent the hard-domines and the black parts are the soft-domines,
Isoplast® 2530 shows an amorphous structure. Due to the chemical compatibility of
TPUs, interdiffusion between Isoplast® 2530 and TPU B layer is expected and the crosssection of the bi-layer TPU film shown in Figure 4.3, in which a interphase with 2micron width is observed [37].
AFM phase images of nominal 65-layer TPU films (in Figure 4.4) reveal that
good multilayered structures are achieved for both extruded and stretched films. The
composition or layer thickness ratio is not 50/50 as expected because: a) despite the use
of a balanced multiplier die and external lubricants there is still some residual
encapsulation of the high-viscosity Isoplast® 2530 by the low-viscosity TPU B; b) The
9-layer feedblock used in this study yields different flow rates in the Isoplast® 2530 and
the TPU B channels, because the total flow rates are the same and the former flows
through five channels, while the latter flows through four. However, taking into
consideration that the viscoelastic ratio of two TPUs is around 100, the multilayered
structure created is a notable success.
4.3.3 Gas barrier properties
Gas (oxygen) permeabilities of mono and multilayer TPU films are shown in
Figures 4.5 and 4.6. As expected, the rigid Isoplast® 2530 has superior barrier properties
than TPU B due to the higher percentage of hard segments. While the bilayer film
exhibits a permeability in-between the two TPUs, the nominal 65-layer film shows even
lower permeability than Isoplast® 2530. The multilayer films were further uniaxially
132
stretched to different strains at 100 oC. The stretched multilayer films show a decrease in
permeability as the deformation increases up to 75%, but a decrease afterwards, up to the
maximum stretch of 300%. Thus, the best gas barrier properties of 0.044 barrer were
found for the 65-layer film with 75% stretch at this temperature. This represents an
almost three-fold improvement relatively to the bilayer film, which has permeability of
0.12 barrer, and a five-fold improvement relatively to TPU B, which has a permeability
of 0.235 barrer. Interestingly these variations of gas permeability with stretch are not
observed if the monolayer Isoplast® 2530 film is uniaxially stretched. The explanation
for this phenonmena will be discussed in next paragraphs.
4.3.4 Thermal properties
The thermal properties of TPU films measured by DSC can be used to shed light
into the observed tendency in oxygen permeability. In Figures 4.7 and 4.8, there are three
pronounced endothermal peaks at 165 oC, 175 oC, and 195 oC, representing dissociation
of long-range order in TPU B, fragmented hard-segment domains (see below) and main
hard-segment domains in Isoplast® 2530, respectively [5-7, 38, 39]. While both
monolayer and bilayer TPU films show almost no endothermal peak around 195 oC, coextruded and stretched (up to 75%) multilayer TPU films exhibit large peaks at this
temperature. Further stretching from 100% to 300% on the multilayer films makes the
endothermal peak shift from 195 oC to 175 oC.
When 75% stretched monolayer
Isoplast® 2530 film and multilayer film are compared, the former only shows a tiny
bump around 195 oC.
Based on the permeability and thermal results of TPU films, the mechanism of
enhanced barrier properties is proposed as shown in Figure 4.9. When mono or bilayer
133
TPU film are extruded, the TPU chains, especially in Isoplast® 2530, with isotropic and
bucky structure due to the thick individual layer thickness are not able to form
impermeable “hard-segments domains” connected by intermolecular hydrogen bondings
between “–NH” and “C=O” groups along the backbone [38, 39]. Once the TPUs are
layered into multilayer structure with individual layer thickness around 6 microns, under
micro-confinement the polymer chains are oriented through the simultaneous squeezing
and spreading process happened in multiplier dies, and start constructing “hard-segments
domains”. The multilayer films are then stretched to 75% to further orient the polymer
chains building up even more and denser “hard-segments domains”, which reduces the
permeability by making a tortuous path for oxygen molecules to pass through the films.
On the other hand, monolayer Isoplast® 2530 with the same 75% stretching is not able to
form such “hard-segments domains”, and reduce the gas permeability, due to the lack of
oriented polymer chains under micro-confinement. Stretching the multilayer TPU films
from 100% to 300% destroys the “hard-segments domains”, instead of producing more
densly packed structures, and micro-crazes start to appear, which causes the gas
permeabilities increase [38].
4.3.5 WAXS
In order to confirm the structural model, WAXS measurements in both the normal
and transverse directions were performed on extruded and stretched TPU films. 2-D
WAXS patterns of monolayer Isoplast® 2530, TPU B and multilayer films are provided
in Figure 4.10. In this, all TPU film show a broad amorphous halo and a sharp diffraction
ring. In TPU B where the long-range order is observed by AFM and DSC, the sharp ring
represents the “hard-segments domains” that have a d-spacing of 7.38 Å, which is in
134
agreement with the results reported by Yunxin et al [7]. As mentioned before, the long
length and high ratio of soft segments facilitate microphase seperation and the formation
of ordered “hard-segments domains”. Compared to TPU B, the sharp ring in Isoplast®
2530 appears at a place closer to the center meaning a slightly larger d-spacing (=8.84 Å)
between the hard-segments sheets because it has no soft segments.
While the extruded multilayer film (Figure 4.10c) exhibit the overlapped
diffraction rings consisting of those in Isoplast® 2530 and TPU B, the multilayer film
with 75% stretch (Figure 4.10d) shows broad relection arcs at the meridian in the center
of the pattern presenting the oriented polymer chains. This tend can be more clearly seen
in 1-D WAXS profile as shown in Figure 4.11, in which the multilayer film shows a
broad peak between the ones of two controlled TPU films at 2Ɵ of 10 and 12 degrees.
This peak becomes broader and more pronounced when the multilayer film is 75%
stretched. The reason is attributed to the fact that the stretching under micro-confinement
helps the polymer chains further oriented forming more and denser “hard-segments
domains”. When the multilayer film is stretched beyond 75% (i.e. to 300%), even though
the arcs in WAXS pattern become sharper and sharper, a new pair of arcs at 2Ɵ of 4.8
degrees appears due to the fragmented “hard-segments domains” (Figure 4. 10e).
4.3.6 Gas barrier properties dependence on stretching temperature
The multilayer TPU films were stretched at different temperatures to examine the
gas barrier properties dependence on stretching temperature. As can be seen in Figure
4.12, with the same amount of stretching, i.e. 75%, the permeability decreases as the
temperature increases up to 120 oC. It is at this stretching temperature that the lowest
permeability of 0.028 barrer is achieved, which is an eight-fold improvement relatively to
135
TPU B, four-fold relatively to the bilayer film, and three-fold relatively to Isoplast® 2530.
However, the permeability increases when the temperature is elevated to 130 oC. As
mentioned before, the enhancement of gas barrier properties is due to the impermeable
“hard-segments domains” that are formed by intermolecular hydrogen bonds. When the
temperature is elevated from 100 oC to 120 oC, the mobility of TPU chains increases, and
the possibility to form “hard-segments domains” becomes higher. This can be proved by
DSC as shown in Figure 4.13 and Table 4.1; the integrated endothermic peak of “hardsegments domains” increases from 9.98 J/g to 12.41 J/g between 100 oC and 120 oC. On
the other hand, at 130 oC the “hard-segments domains” are destroyed due to the high
energy that deconstructs the intermolecular hydrogen bonds, which shows is in agreement
with the findings of Coleman et al [39]. Finally, Figure 4.14 shows the overall
dependency of permeability of multilayer TPU films on temperature and stretching ratio,
and the set of conditions that lead to the lowest permeability is 75% stretching at 120 oC.
4.3.7 Mechanical properties
The effect of micro-confinement on mechanical properties is also pronounced.
The stress-strain curve and the tensile properties for mono-, bi-, and multilayer TPU films
at room temperature are summarized in Figure 4.15 and Table 4.2. As expected, the rigid
Isoplast® 2530 has a high yield stress and a low elongation at break of 95.69%, while the
TPU B shows a ductile mechanical behavior. The bilayer, like the gas barrier properties,
posseses mechanical properties in-between those of the two components. However, the
multilayer film not only shows a high yield stress of 29.17 MPa, but also a high
elongation at break that is closed to 195%, a strong and ductile behavior with toughness
about 76.07 MPa, and the highest fracture stress at 60.49 MPa. These improvements are
136
due to the fact that the stress relief based on the interaction of the microcraks in Isoplast®
2530 with TPU B layers is able to alleviate crazing [27, 40].
4.4 Conclusions
Extruded and stretched nominal 65-layer rigid/soft TPU films were successfully
produced through a novel layer-multiplying co-extrusion and uniaxial stretching and the
dependence of gas permeability on deformation and temperature was studied. Uniaxial
stretching under micro-confinement has a significant effect on gas barrier properties of
TPUs, with up to an eight-fold improvement of 75% stretched multilayer film when
compared to the component polymers. Such a reduction in gas permeability is due to
oriented “hard-segments domains” in the Isoplast® 2530 layers resulting from
intermolecular hydrogen bonds connected between “-NH” and “C=O” groups along the
TPU B backbones. This changes the gas-diffusion path into a tortuous way. Multilayer
TPU films with stretching beyond 75%, instead of further improving the gas barrier
properties, increase the permeability because the “hard-segments domains” start to break
up under these circumstances. Finally, the micro-confinement also shows effect on
mechanical properties, with 100% improvement on elongation at break being found from
monolayer to multilayer TPU film, because the interaction of the microcraks in Isoplast®
2530 with TPU B layers is able to alleviate crazing.
137
4.5 References
1. Furukawa, M.; Mitsui, Y.; Fukumaru, T.; Kojio, K. Polymer, 46, 10817, (2005)
2. Kojio, K.; Kugumiya, S.; Uchiba, Y.; Nishino, Y.; Furukawa, M. Polym. J., 41, 118
(2009)
3. Tang, D.; Macosko, C. W.; Hillmyer, M. A. Polym. Chem., 5, 3231 (2014)
4. Chattopadhyay, D. K.; Raju, K. V. S. N. Prog. Polym. Sci., 32, 352 (2007)
5. Chen, T. K.; Shieh, T. S.; Chui, J. Y. Macromolecules., 31, 1312, (1998)
6. Bajsic, E. G.; Rek, V.; Sendijarevic, A.; Sendijarevic, V.; Frisch K. C. J. Elastom.
Plast., 32, 162 (2000)
7. Wang, Y.; Gupta, M.; Schiraldi, D. A. J. Polym. Sci. Part B: Polym. Phys., 50, 681,
(2012).
8. Gisselfalt, K.; Helgee, B. Macromol. Mater. Eng., 288, 265, (2003)
9. Mishra, A.; Maiti, P. J. Appl. Polym. Sci., 120, 3546 (2011)
10. Prisacariu, C. Polyurethane Elastomers : From Morphology to Mechanical Aspects.
Springer: New York, (2011)
11. Tseng, J.-K.; Tang, S.; Zhou, Z.; Mackey, M.; Carr, J. M.; Mu, R.; Flandin, L.;
Schuele, D. E.; Baer, E.; Zhu, L. Polymer, 55, 8 (2014)
12. Lin, Y.; Hiltner, A.; Baer, E. Polymer, 51, 5807 (2010)
13. Zhang, G.; Lee, P. C.; Jenkins, S.; Dooley, J.; Baer, E. Polymer, 55, 663 (2014)
14. Cheng, W.; Gomopoulos, N.; Fytas, G.; Gorishnyy, T.; Walish, J.; Thomas, E. L.;
Hiltner, A.; Baer, E. Nano Lett., 8, 1423 (2008)
15. Ji S, Yin K, Mackey M, Brister A, Ponting M, Baer E. Opt Eng., 52, 112105 (2013)
16. Lai, C.-Y.; Ponting, M. T.; Baer, E. Polymer, 53, 1393 (2012)
138
17. Singer, K. D.; Kazmierczak, T.; Lott, J.; Song, H.; Wu, Y.; Andrews, J.; Baer, E.;
Hiltner, A.; Weder, C. Opt. Express., 16, 10358 (2008)
18. Ryan, C.; Christenson, C. W.; Valle, B.; Saini, A.; Lott, J.; Johnson, J.; Schiraldi, D.;
Weder, C.; Baer, E.; Singer, K. D.; Shan, J. Adv. Mater., 24, 5222 (2012)
19. Lott, J.; Ryan, C.; Valle, B.; Johnson, J. R.; Schiraldi, D. A.; Shan, J.; Singer, K. D.;
Weder, C. Adv. Mater., 23, 2425 (2011)
20. Wang, H.; Keum. J. K.; Hiltner A.; Baer, E. Macromolecules., 42, 7055 (2009)
21. Zhang, G.; Baer, E.; Hiltner, A. Polymer, 54, 4298 (2013)
22. Langhe, D. S.; Hiltner, A.; Baer, E. Polymer, 52, 5879 (2011)
23. Wang, H.; Keum, J. K.; Hiltner, A.; Baer, E.; Freeman, B.; Rozanski, A.; Galeski, A.
Science, 323, 757 (2009)
24. Carr, J. M.; Mackey, M.; Flandin, L.; Schuele, D.; Zhu, L.; Baer, E. J. Polym. Sci.
Part B: Polym. Phys., 51, 882 (2013)
25. Burt, T. M.; Keum, J.; Hiltner, A.; Baer, E.; Korley, L. T. J. ACS Appl. Mater.
Interfaces., 3, 4804 (2011)
26. Sheng, D.; Tan, J.; Liu, X.; Wang, P.; Yang, Y. J. Mater. Sci., 46, 6508 (2011)
27. Lai, C.-Y.; Hiltner, A.; Baer, E.; Korley, L. T. J. ACS Appl. Mater. & Interfaces, 4,
2218 (2012)
28. Carr, J. M.; Langhe, D. S.; Ponting, M. T.; Hiltner, A.; Baer, E. J. Mater. Res., 27,
1326 (2012)
29. Carr, J. M.; Mackey, M.; Flandin, L.; Hiltner, A.; Baer, E. Polymer, 54, 1679 (2013)
30. Koerner, H.; Kelley, J. J.; Vaia, R. A. Macromolecules, 41, 4709 (2008)
31. Shibayama, M.; Inoue, M.; Yamamoto, T.; Nomura,S. Polymer., 31, 749 (1990)
139
32. Huang, R.; Patz, J.; Silva, J.; Maia, J. M.; Huntington, B. A.; Bonnecaze, R. T.; Cox,
M. ANTEC 2014 Proceedings. April 27-30, Las Vegas, accepted
33. Huang, R.; Patz, J.; Silva, J.; Andrade, R.; Harris, P.; Yin, K.; Huntington, B.;
Bonnecaze, R.; Cox, M.; Maia, J. M. Int. Polym. Proc. submitted
34. Barroso, V.; Covas, J. A.; Maia, J. M. Rheol. Acta, 41, 154 (2002)
35. Harris, P. J.; Patz, J.; Huntington, B. A.; Bonnecaze, R. T.; Meltzer, D.; Maia, J.
Polym. Eng. & Sci., 54, 636 (2014)
36. Silva, J.; Andrade, R.; Huang, R.; Liu, J.; Meltzer, D.; Cox, M.; Maia, J. M. J nonNewtonian Fluid Mech. 2014, submitted.
37. Silva, J.; Maia, J. M.; Huang, R.; Meltzer, D.; Cox, M.; Andrade, R. Rheol. Acta., 51,
947 (2012)
38. Sakurai, S.; Yoshida, H.; Hashimoto, F.; Shibaya, M.; Ishihara, H.; Yoshihara, N.;
Nishitsuji, S.; Takenaka, M. Polymer, 50, 1566 (2009)
39. Coleman, M.M.; Lee, K. H.; Skrovanek, D. J.; Painter, P. C. Macromolecules., 19,
2149 (1986)
40. Ponting, M.; Burt, T. M.; Korley, L. T. J.; Andrews, J.; Hiltner, A.; Baer, E. Ind. &
Eng. Chem. Res., 49, 12111 (2010)
140
Table 4.1: Comparison of integrated endothermic peak of “hard-segments domains” in nominal
65-layer film with 75% stretching at different temperatures.
Temperature
Integrated peak (J/g)
100oC
9.98
110oC
12.20
120oC
12.41
130oC
5.9
141
Table 4.2: Mechanical properties of extruded TPU films.
Elastic modulus
Elongation at
Fracture stress
Yield stress
Toughness
(MPa)
break (%)
(MPa)
(MPa)
(MJ/m3)
Isoplast® 2530
1867.24±98.46
95.69±9.81
46.63±3.81
35.02±3.15
32.02±6.09
TPU B
109.61±3.19
397.86±6.25
41.64±2.23
2.18±0.12
89.61±5.04
Bi-layer
1032.75±100.34
127.40±15.38
34.26±6.86
20.51±2.13
36.65±11.40
65 layers
1265.13±2.38
193.25±1.64
60.49±0.19
29.17±0.71
76.07±2.36
142
Figure 4.1: Rheological results of TPUs: (a) steady shear mode and (b) oscillation shear mode; (c)
extensional rheology for Isoplast® 2530, and (d) TPU B.
143
Figure 4.2: AFM phase images of TPU B (left), and Isoplast® 2530 (right).
144
Figure 4.3: AFM phase images of bilayer Isoplast® 2530/TPU B: low magnification (left), high
magnification (right).
145
Figure 4.4: Morphologies of nominal 65-layer Isoplast® 2530/TPU B films: (a) OM picture of
film as extruded; AFM phase images of (b) extruded, (c) 75% stretched, and (d) 300% stretched
films.
146
Figure 4.5: Oxygen permeability of extruded and stretched TPU films.
147
Figure 4.6: Oxygen permeability of 75% stretched TPU films.
148
Figure 4.7: DSC results of extruded and stretched TPU films (heating rate=10 oC/min-1).
149
Figure 4.8: DSC results of 75% stretched TPU films (heating rate=10 oC/min-1).
150
Figure 4.9: Schematic illustration for micro-confinement effect on forming “hard-segments
domain” in Isoplast® 2530 layer, and microscopic fracture at very high deformation, in which
yellow layer is Isoplast® 2530 and dark blue layer is TPU B, orange boxes are hard segments and
light blue spots are chain extenders.
151
(a)
(b)
(c)
152
(d)
(e)
Figure 4.10: Normal direction 2-D WAXS patterns for various TPU films: TPU B (a), Isoplast®
2530 (b), 65-layer (c), 65-layer with 75% stretch (d), nominal 65-layer with 300% stretch (e).
153
Figure 4.11: 1-D WAXS profiles of various TPU films.
154
Figure 4.12: Oxygen permeability of 75% stretched nominal 65-layer film as function of
temperatures.
155
Figure 4.13: DSC results of nominal 65-layer film with 75% stretching at different temperatures.
156
Figure 4.14: 3-D profile on oxygen permeability of nominal 65-layer film depending on
stretching ratio and temperatures.
157
(a)
(b)
Figure 4.15: Stress-strain curves for various TPU films: (a) full scale, (b) initial area zoomed in.
158
PART II: Twin-screw Compounding Process for Thermoplastic Elastomer
159
CHAPTER 5
Understanding the Distribution and Dispersion of Mineral Oil in
Polypropylene/Styrene-Ethylene-Butadiene-Styrene Blends Upon Compounding
NOTE: Parts of this work have been submitted or published in
“Huang, R.; Chari, P.; Klettlinger, N.; Ling, G.; Tseng, J-K.; Maia, J. M. Int. Polym.
Proc., submitted”
Abstract
This chapter presents a study on the effects of batch and continuous mixing
methods of a three component thermoplastic elastomer, TPE, system of mineral oil
polypropylene (PP) and Styrene-Ethylene-Butadiene-Styrene (SEBS). The experimental
approach keeps the ratios of each TPE component constant, placing emphasis on
distribution and dispersion of oil in the two polymer matrices. Through this procedure,
the effect of different feeding times on glass transition temperature (Tg) of both the PP
and SEBS phases, and the complete TPE system are highlighted. It was observed the
well-known affinity of the oil to SEBS phase when compared to PP, which increases the
glass-transition temperature, Tg, of the blend and leads to processing defects. Also,
increasing residence times of the different phases improves oil absorption in PP, severely
reduces the presence of gels, and lowers the Tg of the batch mixed TPE system. These
results were then confirmed in twin-screw compounding, where we observed a marked
improvement in extrudate appearance and decrease in Tg by adding a second
compounding step.
5.1 Introduction
160
Thermoplastic elastomers that exhibit both plastic and rubber characteristics are
diverse in applications and exist in many forms, such as styrenic block copolymers, and
polyolefin blends [1-6]. Applications of thermoplastic elastomers range from filling the
soles of shoes to cables in headphones [7-9]. With many different classes of TPE and
range in applications, finding favorable characteristics to strengthen the material will
benefit a wide array of products. An important complex included from these classes is
the blend of polypropylene (PP) and Styrene-Ethylene-Butylene-Styrene (SEBS), which
possesses favorable tensile modulus and strength [10-12]. With PP droplets imbedded in
the SEBS matrix, the blend becomes stiffer, more easily processed and cost-effective, and
tougher since the SEBS helps prevent the potential crazing of PP at large deformations
[13-17].
There are a variety of processing methods to produce styrenic TPE for its many
different applications. These include extrusion, injection molding, compression molding,
and sometimes blow molding and heat welding [18-23]. Twin-screw extrusion is able to
provide the TPE system with the high dispersive and distributive mixing required for
proper blending of PP and SEBS. However, due to the high elasticity and viscous
dissipating heat of SEBS in the molten state, it is difficult to process this material. Thus,
normally a plasticizer such as mineral oil is added to the formulation. At optimal amounts
the oil becomes more and more of an extender, which will soften the polymer matrix [2426]. In the meanwhile, due to the polarity difference, the distribution of oil in each PP
and SEBS phase plays an important role in determining the mechanical and optical
properties of the TPE. In addition to the elongation extension, the addition of a plasticizer
161
also reduces the glass transition of the polymers, allowing the rubbery region of the
material to be reached at a faster rate [27].
Even though it has been shown before that increasing the content of oil leads to a
decrease in Tg of PP/SEBS blends, the distribution and dispersion kinetics of oil in the
polymeric phases is still not well understood [28-30]. In this paper, we begin by studying
thermo-rheologically the mixing kinetics of the oil in each of the polymeric phases
separately, in an internal batch mixer. Then we repeat the process for the complete TPE,
varying the feeding protocol and mixing time in order to minimize the Tg of the final
blend. Finally, we upscale the process for twin-screw extrusion. We subjected TPE to one
and two passes, and were able to confirm not only the mixing kinetics but also minimize
both the Tg of the blend and the presence of gels in the final extruded product with the
second extrusion cycle.
5.2. Experimental and method
5.2.1 Materials
All the materials used in this work were supplied by Saint-Gobain Performance
Plastics. The isotatic polypropylene has a density of ρ=0.9 g/cc as measured by ASTM
D1238 and a melt flow rate of 4.2 g/10min. The styrenic block copolymer was an unhydrogentated styrene-ethylene-butadiene-styrene (SEBS) with a hardness of Shore A
60. The mineral oil, which serves as a plasticizer, has a density of ρ=0.88 g/cc
5.2.2 Internal Batch Mixer
162
The PP/SEBS/oil blends, at a constant ratio of 20/35/45 percentage by weight,
typical of industrial applications, were prepared in an internal batch mixer (HAAKE™
Rheomix OS Lab Mixers with CAM blades/rotors) at 175 oC and 250 rpm, in different
feeding orders. In particular, PP/oil at ratio of 8:1 by weight was first fed into the
chamber of mixer, followed by different amounts of oil only, and finally and then the rest
of the oil and SEBS/oil at ratio of 1:1 were added. The mixing time of each of these three
steps was 75 seconds, and the specifics were shown in the Table 1. After the mixing step,
the samples were hot pressed by compression mold (Carver hydraulic compression
molder) to make thin films with thickness less than 25 microns at 175 oC.
5.2.3 Twin-screw Compounding
PP/SEBS/oil at a constant ratio of 20/35/45 percentage by weight were first dry
blended with a KitchenAid Professional 600 Series 6-Quart Stand Mixer for 2 hours at
room temperature, and then the mixture was melt compounded in a 24 mm intermeshing
co-rotating twin-screw extruder (Thermo Scientific™ TSE 24 MC, L/D=40/1). The screw
configuration consists of 5 conveying blocks and 4 kneading blocks, as shown in table 2.
The processing temperature was set at 175 oC, and a screw speed of 400 rpm and
throughput of 30 lbs/hour were used for each compounding process. After the materials
were extruded from a 3-hole strand die, they were quenched through a water tank and
diced by a pelletizer. Both the first and second compounding passes were performed
under these processing conditions.
5.2.4 Characterization
163
Oscillation shear experiments of the blends were performed via a rotational
rheometer (Thermo Fisher MARS III) with 20 mm parallel plates system. A constant
shear stress of 100 Pa (within the linear viscoelastic regime) and temperature of 175 oC
were used for the frequency sweeps from 100 to 0.01 Hz. Dynamic moduli and complex
viscosity were collected to analyze the effect of different compounding processes on oil
distribution and dispersion in PP/SEBS.
For Atomic Force Microscopy (AFM), the samples were embedded and fixed in
epoxy at room temperature for 1 day. A Leica microsystmes EM FC6 ultramicrotome
(Buffalo Grove, IL) was applied to cut the cross sections of the sample at -90 oC
transveral to the extrusion direction. The samples were examined by a Digital
Laboratories Nanoscope IIIa AFM (Digital Instruments, Santa Barbara, CA) operating in
tapping mode at room temperature. AFM phase images of the blends with different
compounding processes were analyzed via the NanoScope software to obtain modulus
differences and morphology information.
The glass transition temperatures (Tg) of the blends were determined by Dynamic
Mechanical Analysis (DMA). The experiments were performed by using a TA
Instruments DMAQ800 within a liquid nitrogen atmosphere and with heating and cooling
rate of 3 oC/min. Multi-Frequency-Strain mode at frequency of 1 Hz was applied to all
the DMA measurements. The peak temperatures in loss modulus were used to determine
the Tg of each sample.
Broadband Dielectric Spectroscopy (BDS) measurements were conducted by an
Alpha–A broadband dielectric analyzer, Novocontrol. The tested samples were coated
with round gold electrodes on both sides (diameter = 1 cm, thickness = 20 nm). The
164
temperature of the sample was controlled by Novocontrol Quatro Cryosystem.
Temperature swipe BDS measurement was carried out with the temperature ramp rate of
2 °C/min.
5.3 Results and discussion
The first part of the work is concerned with establishing the oil absorption
capability of each of the polymeric phases. As seen in Figure 5.1, even for PP/oil ratios
as low as 8/1, only a small amount of oil can be absorbed by PP at room temperature,
with most staying outside the PP pellets. In contrast, the oil is completely absorbed by
SEBS even at the ratio of 1:1. This is majorly attributed to the polarity and free volume
difference of between PP and SEBS.
The maximum ratio of PP/oil that can be dry-blended prior to feeding in the batch mixer
of the extruder is approximately 8:1. Beyond this ratio, the extra oil stays outside the PP pellets at
room temperature no matter the duration of the dry blending process. This ratio of PP and oil is
first fed into the mixing chamber. The second step is the feeding of extra oil only and finally the
rest of the oil and the SEBS are fed at a ratio of 1/1. In order to study the effect of different
mixing times, two cycle times of 15-45-15 (seconds) and 15-30-30 (seconds) were imposed in
each step. As seen in Figure 5.2, the film with less mixing time in SEBS/oil exhibits white
impurities (possible gels), while the other material is totally transparent. This might be attributed
to the fact that the time for transferring oil from SEBS phase into PP phase is too short to
eliminate the inhomogeneous/unplasticized PP, which causes gel impurities. This is confirmed by
DMA, as shown in Figure 5.3. The inhomogeneous/unplasticized PP part in the mixture of 1545-15 (sec) shows a higher Tg than that of 15-30-30 (sec).
PP/oil (8:1), SEBS/oil (8:1), and SEBS/oil (1:1) were then compounded twice in
the twin-screw extruder. DMA shows that for PP/oil (8:1), Tg decreases from 30 oC to 165
40 oC between the first and second passes in the extruder, as shown in Figure 5.4(a).
This means the oil is more homogeneously distributed and dispersed in PP after the
second pass. Figure 5.4(b) and (c) show the DMA results of SEBS/oil systems with ratio
of 8:1 and 1:1 respectively. Even though the Tg of SEBS/oil at ratio of 1:1 (-60 oC) is
10oC lower than that of 8:1 (-50 oC) due to the plasticizing effect from oil, in both cases
there is no shift in Tg between the first and second passes. This means the oil can be
easily distributed and dispersed in the SEBS.
Therefore, when PP/SEBS/oil are dry blended at room temperature, it is to be
expected that most of the oil be absorbed into the SEBS phase. This may cause an
inhomogeneous oil distribution in the extruded TPE blend if one compounding step is not
sufficient to transfer some of the excess oil in the SEBS phase to the PP phase and
distribute equally throughout both. Therefore, the TPE dry blend was also subject to two
compounding passes in the extruder.
Figure 5.5 shows the DMA results on both extruded samples and it is possible to
observe that: a) There is a decline in loss modulus peak between the first and second pass
samples as a function of temperature; and b) There is a transition from two glass
transition peaks for the one-pass blend to only one in the two-pass blend. The peak at
approximately -65 oC is the Tg of the SEBS phase, which is consistent with literature
results for ~50 wt% oil in SEBS [28]. The shoulder that can be observed at -45 oC is the
Tg of the highly plasticized PP phase. This is a significantly lower value of Tg than that
of PP alone, but is again in agreement with previous literature findings for highly
plasticized PP [28]. Interestingly, after the second compounding these two Tg peaks
merge at -57 oC instead of moving towards lower temperatures. This is a strong
166
indication that the second pass is indeed inducing a transfer of oil from the SEBS to the
PP phase, thus increasing the Tg of the former. It is also an indication of the more
homogeneous distribution and dispersion of oil in the TPE. Sengers pointed out that oil
actually acts as a plasticizer, softening the material as its concentration increases and that,
surprisingly, the effect is much more pronounced for PP than for SEBS [28]. In this paper,
even though the oil concentration is constant, better distribution and dispersion of oil will
create larger surface to volume ratio, leading to the same effect as increasing oil
concentration.
These mechanical spectroscopy results are confirmed by BDS, Broadband
Dielectric Spectroscopy. Since the dielectric response of the molecule depends on the
changes in the surrounding viscosity, BDS monitors the mechanical glass transition
evolution during the compounding process [26]. Figure 5.6 shows that there are two
peaks of loss permittivity (the equivalent of the mechanical loss modulus) for the first
pass sample, and only one peak for the second pass. Interestingly, the high Tg peak in the
first pass occurs at different temperatures in DMA and BDS experiments. The reason for
this discrepancy is not clear.
Even though DMA and BDS results show significant differences in the Tg peaks
after one and two passes, rotational rheometry data shows that rheologically the two
materials have quantitatively the same behavior, see Figure 5.7. In both cases the
behavior is typical of elastomers, with moduli showing a very weak dependence on
frequency and the storage modulus being much higher than the loss modulus at all
frequencies.
167
Atomic Force Microscopy, AFM, was used to investigate the morphology of the
blends and the similar rheological behavior between the first and second passes. Figure
5.8(a) and (b) show the phase images of first and second pass samples; the bright areas
represent the PP phase and the dark areas represent the SEBS phase. As expected from
the rheometry results, a similar morphology is observed for these two samples, therefore
confirming that the diffences observed by DMA are due to oil dispersion (which is not
visible by AFM) and not to different bulk morphologies (which would be).
5.4 Conclusions
Blends of PP/SEBS/oil were successfully prepared by batch mixing and twinscrew compounding. Batch mixing results show that both PP and SEBS are highly
plasticized by the oil, and different feeding times would change the optical properties
(transparency) and Tg of the film. Even though PP acts as oil resistant material and SEBS
is highly affiliated with oil during dry blending at room temperature, PP can be more
plasticized by oil upon melt compounding. Extrusion results show that due to the higher
affinity of the oil with the SEBS phase, dry blending results in an oil-depleted PP phase
upon compounding, with both DMA and BDS showing a two Tg system. Therefore, a 2pass twin-screw compounding is necessary to homogeneously distribute the oil in the
TPE system, even if the general morphology, as determined by AFM and rotational
rheometry, does not evolve noticeably between the two passes. Tg at -65 oC from the
SEBS phase and Tg around -45 oC from PP phase merge towards -57 oC in the second
pass. However, evolutions in morphology and rheology are not observed noticeably
within 2-pass compounding.
168
5.5 References
1. Chen, Y., Kushner, A. M., Williams, G. A., Guan, Z., Nat Chem, 4, 467 (2012)
2. Shi, W., Lynd, N. A., Montarnal, D., Luo, Y., Fredrickson, G. H., Kramer, E. J.,
Macromolecules, 47,2037 (2014)
3. George, J., Varughese, K. T., Thomas, S., Polymer, 41, 1507 (2000)
4. Sonnenschein, M. F., Ginzburg, V. V., Schiller, K. S., Wendt, B. L., Polymer, 54, 1350
(2013)
5. Thomas, S., George, A., Eur. Polym. J., 28, 1451 (1992)
6. George, S., Ramamurthy, K., Anand, J., Groeninckx, G., Varughese, K.T., Thomas, S.,
Polymer, 40, 4325 (1999)
7. Kollosche, M., Kofod, G., Appl. Phys. Lett., 96, 071904 (2010)
8. Armstrong, S., Freeman, B., Hiltner, A., Baer, E., Polymer, 53, 1383 (2012)
9. Khan, U., May, P., O’Neill, A., Vilatela, J. J., Windle, A. H., Coleman, J. N., Small, 7,
1579 (2011)
10. Lin, Z., Chen, C., Guan, Z., Tan, S., Zhang, X., Al, L. J., J. Appl. Polym. Sci., 122,
2789 (2011)
11. Panaitescu, D. M., Vuluga, Z., Radovici, C., Nicolae, C., Polym. Test., 31, 355 (2012)
12. Vuluga, Z., Panaitescu, D. M., Radovici, C., Nicolae, C., Iorga, M. D., Polym. Bull.,
69, 1073 (2012)
13. Gupta, A. K., Purwar, S. N. J., J. Appl. Polym. Sci., 29, 1595 (1984)
14. Gupta, A. K., Purwar, S. N. J., J. Appl. Polym. Sci., 29, 3513 (1984)
15. Gupta, A. K., Purwar, S. N. J., J. Appl. Polym. Sci., 31, 535 (1986)
16. Gupta, A. K., Srinivasan, K. R. J., J. Appl. Polym. Sci., 47, 167 (1993)
169
17. Wilkinson, A. N., Laugel, L., Clemens, M. L., Harding, V. M., Marin, M., Polymer,
40, 4971 (1999)
18. Kusmono, Ishak, Z. A. M., Chow, W. S., Takeichi, T., Rochmad., Eur. Polym. J., 44,
1023 (2008)
19. Stribeck, N., Fakirov, S., Macromolecules, 34, 7758 (2001)
20. Elleuch, R., Elleuch, K., Salah, B., Zahouani, H., Mater. Des., 28, 824 (2007)
21. Yu, Z., Lei, M., Ou, Y.; Yang, G., Polymer, 43, 6993 (2002)
22. Saikrasun, S., Amornsakchai, T., Sirisinha, C., Meesiri, W., Bualek-Limcharoen, S.,
Polymer, 40, 6437 (1999)
23. Li, Y., Oono, Y., Kadowaki, Y., Inoue, T., Nakayama, K., Shimizu, H.,
Macromolecules, 39, 4195 (2006)
24. Chang, K., Robertson, M. L., Hillmyer, M. A., ACS Appl. Mater. Interfaces, 1, 2390
(2009)
25. Ohlsson, B., Tornell, B., Polym. Eng. Sci., 36, 1547 (1996)
26. Sengers, W. G. F., Sengupta, P., Noordermeer, J. W. M., Picken, S. J., Gotsis, A. D.,
Polymer, 45, 8881 (2004)
27. Jha, A., Dutta, B., Bhowmick, A. K. J., J. Appl. Polym. Sci., 74, 1490 (1999)
28. Sengers, W. G. F., Wübbenhorst, M., Picken, S. J., Gotsis, A. D., Polymer. 46, 6391
(2005)
29. Sengupta, P., Noordermeer, J. W. M., Polymer, 46, 12298 (2005)
30. Sengupta, P., Noordermeer, J. W. M., Macromol. Rapid Commun., 26, 542 (2005)
170
Table 5.1: The different feeding time of PP/oil/SEBS in internal batch mixer (unit: second)
Samples
PP/oil (8:1)
Oil
SEBS/oil (1:1)
TPE 15-45-15
15
45
15
TPE 15-30-30
15
30
30
171
Table 5.2: Twin screw configuration: the order starts from feeder to exit die; each conveying
element is 2.4 cm, and each kneading element is 0.6 cm; every 4 kneading elements together have
the same twisting angle in arrangement.
Order
Elements
Twisting angle
Conveying
1
10
n/a
Kneading
2
4×4
30o/60o/60o/90o
Conveying
3
4
n/a
Reversed conveying
4
0.5
n/a
Kneading
5
2×4
60o/60o
Conveying
6
4
n/a
Reversed conveying
7
0.5
n/a
Kneading
8
3×4
60o/60o/90o
Conveying
9
3
n/a
Kneading
10
2×4
60o/60o
Conveying
11
7
n/a
Total
n/a
40
n/a
172
Figure 5.1: Pictures of oil absorption in PP (a-c), SEBS (d-f) phase at room temperature.
173
Figure 5.2: Optical pictures of PP/oil/SEBS with different mixing time: 15-45-15 (left), 15-30-30
(right).
174
Figure 5.3: DMA results of PP/oil/SEBS with different mixing time.
175
(a)
(b)
176
(c)
Figure 5.4: DMA results for extruded PP/oil (8:1) (a), SEBS/oil (8:1) (b), and SEBS/oil (1:1) (c)
blends.
177
Figure 5.5: DMA results for the extruded TPE system after one and two compounding cycles.
178
Figure 5.6: BDS of the 1st and 2nd pass samples.
179
(a)
(b)
Figure 5.7: Oscillation shear rheology of 1st and 2nd pass samples: (a) dynamic moduli as
function of frequency; (b) complex viscosity as function of frequency.
180
(a)
(b)
Figure 5.8: AFM phase images of TPE: (a) sample from 1st pass, (b) sample from 2nd pass.
181
APPENDIX
182
APPENDIX
Interplay Between Rheological and Structural Evolution of Benzoxazine Resins
During Polymerization
NOTE: Parts of this work have been submitted or published in
“Huang, R.; Carson, S.; Jorge, S.; Agag, T.; Ishida, H.; Maia, J. M. Polymer, 54, 1709
(2013)”
A.1 Introduction
Polybenzoxazines are a class of phenolic resins that possess various unusual and
advantageous properties [1]. While traditional phenolics offer many useful properties and
still hold the majority of the share of the thermoset market today, their use also has
various problems when considering synthesis, processing, and end use. Examples of these
include the use of harsh chemicals during synthesis and polymerization, large volumetric
shrinkage during processing, production of water during polymerization which leads to
void formation, and an inherent brittleness of the material in the polymerized and crosslinked form [2]. Polybenzoxazines offer the same advantages of traditional phenolics
such as excellent mechanical strength, thermal and thermo-oxidative stabilities, good
chemical resistance, abrasion resistance, and flame resistance but eliminate most of the
problems associated with them [3-8]. Properties such as near-zero volumetric change
upon polymerization, very high char yield, low water absorption, and a by-product free
polymerization are unique to benzoxazine based resins [3, 9-12]. Polybenzoxazines also
show various forms of hydrogen bonding within the final chemical structure, which
183
contribute greatly to its unique properties [13, 14]. Relatively high polymerization
temperature can be inconvenient for some applications and, as is for all thermosetting
resins, the improvement in toughness may sometimes be needed, although polymeric
benzoxazine precursors, such as main-chain and side-chain type polybenzoxazines that
can be later cross-linked, offer much better toughness characteristics [15-26]. When
compared with traditional thermoset phenolics and epoxies, polybenzoxazines offer a
number of useful advantages that make them an attractive alternative, in addition to
offering new application opportunities using those unique properties mentioned earlier.
The potential for rich molecular design flexibility is one of the most important
properties of a benzoxazine-based resin, since it makes it possible for the resin properties
to be tailored specifically to an application and still offer all of the same chemical and
processing advantages. It is mostly for this reason that polybenzoxazines are very viable
materials that find increasing number of applications, from replacements of traditional
phenolics and epoxies to specifically tailored high performance materials [1].
Chemically, benzoxazine resins are synthesized through a Mannich condensation
of a phenolic derivative, an amine, and formaldehyde. Water is the sole byproduct of this
monomer synthesis. The potential for molecular versatility comes from the wide
availability of phenolic derivatives and primary amine compounds that can be used in the
production of the material. To form a crosslinked thermoset structure, multifunctional
phenolics and amines are used. Perhaps the most common form of benzoxazine resin that
exemplifies all of the resin family’s typical properties is produced from Bisphenol-A and
aniline, commonly designated as BA-a [12, 27-29]. Another common variation of
bifuntional benzoxazines, which is more advanced in heat resistance and electrical
184
insulation than BA-a, is based on methylene dianiline, and is designated as P-ddm and its
cationic ring-opening mechanism (the same as for BA-a) is shown in Figure A.1 [30].
The aim of this study is to understand the rheological evolution of phenol and
methylene dianiline (P-ddm) benzoxazines, which is essential to its successful processing.
The polymerization kinetics of benzoxazines by using differential scanning calorimetry
(DSC) has been reported elsewhere [27, 31-34]. However, by combining this with
rheological means we expect to establish a complete and fundamental understanding of
the kinetics, and monitor how benzoxazine acts over different polymerization
temperatures. This has not been done before. In particular, the polymerization kinetics
was extensively studied using Rheometer, DSC, and Fourier transform infrared
spectroscopy (FT-IR).
A.2 Experimental and method
A.2.1
Preparation
of
benzoxazine
monomer,
bis(3-phenyl-3,4-dihydro-2H-
benzo[e][1,3]oxazine-6-yl)methane (abbreviated as P-ddm)
110g P-ddm was synthesized according to the modified method of the procedure
for the difficult aromatic amines [35]. Some of the possible structures in these assynthesized benzoxazines are shown in Figure A.1. Diaminodiphenylmethane (DDM)
(>99%), paraformaldehyde (96%), aniline, 1,4-dioxane, and a mixture of xylene isomers
were purchased from Aldrich Chemical Company. All chemicals were used without
further purification.
A.2.2 Characterization
185
1
H NMR spectra were acquired in deuterated dimethyl sufoxide on a Varian
Oxford AS600 at a proton frequency of 600 MHz. The average number of transients
for 1H is 64. A relaxation time of 10 s was used for the integrated intensity determination
of 1H NMR spectra.
Rheological analysis was performed for P-ddm (~0.7g) using an Anton Paar
Rheometer (Model Physica MCR 501) and 25mm disposable parallel plates. Small
amplitude oscillatory shear (SAOS) time sweep experiments over temperatures ranging
from 140°C to 220°C in increments of 20°C and temperature sweep from 140°C to 220°C
with heating rate of 3°C/min were performed using a constant frequency of 10 rad/s for
all experiments. During the measurements, the stress was continuously increased between
10 Pa and 600 Pa, so as to maintain an acceptable signal strength, i.e., a high enough
strain to produce consistent and reproducible results, while keeping the materials’
response in the linear viscoelastic regime as polymerization progresses. Stress relaxation
experiments with a shear step strain of 10% were performed on fresh P-ddm samples, as
well as on samples corresponding to critical stages during the polymerization process
(which will be explained later) at 140oC. In these cases the experiment was stopped and
the temperature lowered to 110oC so as to prevent further polymerization within the
testing time scale.
To study the non-isothermal behavior of P-ddm benzoxazines, differential
scanning calorimetric (DSC) analysis was carried out on a TA Instruments Q-100 DSC.
Non-isothermal experiments on monomer at a ramp rate of 3oC/min, or products
collected at certain polymerization stages at 10oC/min were carried out from 0 to 300°C.
The reaction was considered complete when the curves leveled off to the baseline and no
186
more drastic changes in heat were observed. Experiments were always performed below
300°C to prevent any possible degradation inside the chamber. After the exothermic peak,
when the DSC curve reached the baseline level again, the sample was cooled rapidly to
0°C. Further heating of the sample was done to determine the residual heat of reaction.
Fourier transform infrared (FT-IR) spectra were obtained using a Bomem
Michelson MB100 FT-IR spectrometer equipped with a deuterated triglycine sulfate
(DTGS) detector and a dry air purge unit. Co-added spectra of 32 scans were recorded at
a resolution of 4 cm-1. Transmission spectra were obtained at room temperature using the
KBr pellet technique for partially or completely polymerized samples.
A.3 Results and discussion
A.3.1 Synthesis and characterization of P-ddm
H NMR spectrum of P-ddm is shown in Figure A.2 ((CD3)2SO, 600 MHz, δ):
1
6.49-7.39 ppm (Ar-H), 5.33 ppm (O-CH2-N), 4.54 ppm (O-CH2-N), 3.67 ppm (Ar-CH2Ar), 2.47 ppm ((CD3)2SO), 3.30 ppm (H2O), 3.54 ppm (dioxane). The oxazine ring
content in the whole composition can be calculated by the following equation:
Ring content (%) =
I
,× 100
2I
where I is the integrated intensity of the methylene protons of Ar-CH2-N in the
benzoxazine ring, I’ is the integrated intensity of the methylene protons of Ar-CH2-Ar.
The oxazine ring content thus determined for P-ddm is 95%. We did not attempt to
further purify the monomer since this level of purity determined is similar to that of
technical grades of benzoxazine monomers, so it’s a production-representative sample.
187
A.3.2 Rheological properties evolution of benzoxazine during polymerization
The rheological properties of P-ddm are able to provide basic insights about the
processability of polybenzoxazine, since the viscoelastic response of the systems is
highly sensitive to the structural changes during the polymerization. Figures A.3 and A.4
show the evolution of both viscoelastic moduli (the storage modulus, G’, and the loss
modulus, G’’) and tan δ (G’’/G’), clearly revealing the existence of a two transitions
during polymerization, especially at the lower temperatures. As with any chemical
processes, increasing the temperature at which the reaction takes place increases the
polymerization rate, as shown in Figure A.5. For this reason, the first transition is not
observed because it occurs too quickly to be detected experimentally.
Figure A.3 displays all time sweep data for the P-ddm experiments. Investigation
of these curves reveals the unique behavior of benzoxazine polymerization, with all
curves up to 180°C showing two transitions in the polymerization process. This can be
seen even more clearly in Figure A.4, which depicts the time evolution of tan δ. Peaks in
this value indicate that the G” component of the dynamic moduli was at a local relative
maximum compared to the G’ component, and were indicators of structural processes,
such as relaxation or buildup of molecular network. As the test temperature increases so
does the rate of polymerization. Both peaks shift to lower times, while the first peak
eventually disappears completely at 220oC. The disappearance of this first peak at high
temperatures is due to the fact that the process, whatever its nature, is faster than the time
necessary to load the sample and stabilize it at the measurement temperature. Another
characteristic of this peak is that its magnitude decreases greatly as temperature increases.
188
Both these facts point to a transient kinetic process that reduces in magnitude and
increases in velocity as the polymerization temperature increases.
It should also be noted that the elasticity of P-ddm at the beginning of time sweep
at 160oC is too low to be detected by rheometer, while the viscosity keeps increasing.
This might be due to the temporary intermolecular hydrogen-bonding structures forming
between monomers and linear macromolecular chains at 140oC, as can be seen in Figure
A.6(a), which disappear when increasing the temperature from 140oC to 160oC [36, 37].
Such structural transformation can take place while the molecular weight of the resin is
still small and does not contribute significantly to the elasticity build-up.
This is
consistent with the finding that the formation of covalent bonds among monomers lags in
time of the ring opening reactions, although eventually they match in time [36].
These two transitions during polymerization at lower temperatures (i.e. 140oC,
160oC, 180oC) for P-ddm must be related to the gradual buildup of molecular structure
during the entire experiment, with the material going from monomers to oligomers,
unentangled polymer chains and eventually to polybenzoxazine with a fully crosslinked
structure. Thus, we are in condition to put forward the following hypothesis:
i) The first tan δ peak, which was kinetic in nature, probably
corresponds to monomer ring-opening and the opened Mannich base
hydrogen bonds to each other as shown in Figure A.6 (b). There is very little
polymerization taking place at this condition. During this process, the
viscosity grows faster than elasticity, which leads to an increase in tan δ. As
the polymerization reaction starts later and six-membered intramolecular
hydrogen bonds between the hydroxyl group and the nitrogen atom is formed,
189
this behavior was expected to be inverted, with elasticity growing faster than
viscosity, thus leading to a decrease in tan δ. Thus, the first peak in tan δ is
probably due to the onset of an increased intermolecular interaction and
consequent energy dissipation.
ii) The second tan δ peak present in the data probably represents the
actual polymerization occurring during the experiment. This peak changes its
pattern from very distributed over a long period at lower temperatures to very
quick and shifted to shorter times at higher temperatures. This is consistent
with the fact that once the threshold level of chemically ready linear polymer
is reached, a buildup of a large molecular network begins to occur, with
phenolic hydroxyl groups formed from benzoxazine ring-opening serving as
polymerization catalyst. This is an autocatalytic process of benzoxazine resins
as reported in the literature [27]. Thus, the second peak should correspond to
the polymerization process, in which polymer chains propagate and crosslink
with each other and complete the 3-dimensional network.
It is interesting to notice that if one performs a temperature sweep, which is
probably the most common way of studying the temperature–induced polymerization
processes, two tan δ peaks during polymerization are also observed, as can be seen in
Figure A.7. In fact, the continuous increase in temperature destroys the temporary
structure of intermolecular hydrogen bond after ring opening, which is attributed to its
first tan δ peak. The second tan δ peak is due to the actual polymerization process and 3D network formation.
190
In order to test these hypotheses, further rheological, thermal and chemical
experiments were performed. The first of the former was stress relaxation after a step
shear on P-ddm at four different stages of the process: beginning, 1st and 2nd tan δ peaks,
and at the end of the time sweep. Figure A.8 shows the typical results for the relaxation
modulus, G(t). For P-ddm resins at a polymerization temperature of 140oC, it is clear that
in the beginning and at the 1st tan δ peak, the material still behaves like a viscoelastic
liquid, i.e., a non-zero G(t), decaying to zero over time. The large initial drop in the
monomer sample is due to the instrument delay time at the beginning of the experiment,
which arises from the very low viscosity of the monomer. On the contrary, by the time
the 2nd tan δ peak was reached, G(t) no longer relaxes to 0, but does so to a finite value.
This is the behavior typical of a viscoelastic solid and clearly indicates that a permanent
solid-like structure has been formed. The finding clearly reinforces the theory that the
first relaxation mode corresponds to the transition from monomer to unentangled low
molecular weight, intermediates which are loosely connected via hydrogen bonding,
whereas the second relaxation process corresponds to the main crosslinking process.
A.3.3 Thermal properties
Non-isothermal DSC experiments were also performed on benzoxazine as shown
in Figure A.9. The thermograms show that the melting and polymerization temperature
for P-ddm are around 50oC and 213oC, respectively. In order to find if there is any
difference among the polybenzoxazines polymerized at different temperatures, samples
were collected after the polymerization process was rheologically complete, i.e., the
dynamic moduli had reached a plateau, and characterized by non-isothermal DSC. These
results are depicted in Figure A.10 and show the lower the polymerization temperature,
191
the greater the exothermal heat. Therefore, those products polymerized at lower
temperature have lower final degrees of polymerization, even though they achieve similar
values of the dynamic moduli as the ones at higher polymerization temperatures. This
was not expected and is probably due to the limitation of rheometer which was not able to
show the dynamic moduli change once the sample tested was crosslinked and reached a
threshold of dynamic moduli. Additionally, the polybenzoxazine’s Tg shifts to higher
temperatures with higher crosslinking temperatures, which supports the notion of
increased crosslink density with increased polymerization temperature.
A.3.4 Structural evolution of benzoxazine during polymerization
Finally, the polymerization process of P-ddm monomers was also investigated by
FT-IR, monitoring the changes of the characteristic peak of benzoxazine. Figure A.11
represents the vibrational spectra of the benzoxazine monomer, the material at the first
and second peaks of tan δ, and of the final crosslinked material. The peaks centered at
754 and 694 cm-1, corresponding to the mono-substituted benzene, were present at
different stages of the polymerization. However, the characteristic absorption of the
benzene ring to which oxazine ring is attached at 947 cm-1, the band at 1496 cm-1
corresponding to the in-plane carbon-carbon stretching of tri-substituted benzene ring,
and the absorptions at 1030 and 1231 cm-1 corresponding to the symmetric and
asymmetric C-O-C bonds of the benzoxazine all decreased with increasing
polymerization time, indicating the ring-opening reaction of benzoxazine had taken place.
But, there was no detectable evidence of chain propagation at this stage.
On the contrary, a small, heavily overlapped shoulder around 1478 cm-1 appears
from the second tan δ peak onward. This corresponds to the in-plane carbon-carbon
192
stretching of the tetra-substituted benzene ring, as the methylene bridges form in the free
ortho-positions of the phenolic structures, and is attributed to the polymerization [38].
Figure A.12 shows IR spectra of the final resins at all polymerization
temperatures and after a temperature sweep. In it the peak of benzoxazine-related band at
947 cm-1 diminishes, while the peak at 1496 cm-1 gradually shifts to 1478 cm-1 with
increasing polymerization temperature, indicating that benzoxazines polymerized at
higher temperature have a higher degree of polymerization [39].
A.4 Conclusions
P-ddm based benzoxazines polymerized at different temperatures was studied in
terms of its chemo-rheological and thermal behavior. Even though benzoxazine
polymerizes faster at higher polymerization temperatures, the same ultimate dynamic
moduli and tan δ were achieved for all cases. Rheological results of P-ddm at different
polymerization temperatures also clearly show there are two transitions during
polymerization at lower polymerization temperatures, and one of them at elevated
polymerization temperatures. According to the stress relaxation and FT-IR results, the
first tan δ peak was assigned to the ring-opening reaction and subsequent hydrogen bond
formation of open Mannich base of benzoxazine, while the second tan δ peak is the chain
propagation and buildup of a large molecular network. FT-IR and DSC results of
benzoxazines reaching the plateau of dynamic moduli at higher polymerization
temperature show a higher degree of polymerization with increasing temperature. The
reason for the absence of the first peak at elevated temperatures is purely due to its
kinetics, which is very fast and is complete before the temperature in the rheometer can
be stabilized.
193
A.5 References
[1]
Ishida H. in "Handbook of Benzoxazine Resins," Ishida H and Agag T, Eds.
Elsevier, Amsterdam, pp. 3-81 (2011).
[2]
Pilato L. "Phenolic Resins: A Century of Progress." Springer, New York (2010).
[3]
Ishida, H.; Allen, D. J. Journal of Polymer Science Part B-Polymer Physics, 34,
1019 (1996).
[4]
Low, H. Y.; Ishida, H. Journal of Polymer Science Part B-Polymer Physics, 36,
1935 (1998).
[5]
Low, H. Y.; Ishida, H. Journal of Polymer Science Part B-Polymer Physics, 37,
647 (1999).
[6]
Kim, H. D.; Ishida, H. Journal of Applied Polymer Science, 79, 1207 (2001).
[7]
Wu, Y. Q.; Zeng, M.; Jin, H. Y.; Xu, Q. Y.; Lu, L. Y.; Wang, J.; Hou, S. E.
Tribology Letters, 47, 389 (2012).
[8]
Espinosa, M. A.; Galia, M.; Cadiz, V. Polymer, 45, 6103 (2004).
[9]
Ishida, H.; Low, H. Y. Macromolecules, 30, 1099 (1997).
[10]
Liu, X.; Gu, Y. Acta Polymerica Sinica, 5, 612 (2000).
[11]
Kim, H. J.; Brunovska, Z.; Ishida, H. Polymer, 40, 1815 (1999).
[12]
Ning, X.; Ishida, H. Journal of Polymer Science Part a-Polymer Chemistry, 32,
1121 (1994).
[13]
Wirasate, S.; Dhumrongvaraporn, S.; Allen, D. J.; Ishida, H. Abstracts of Papers
of the American Chemical Society, 213, 81-Pmse (1997).
[14]
Kim, H. D.; Ishida, H. Journal of Physical Chemistry A, 106, 3271 (2002).
194
[15]
Kimura, H.; Matsumoto, A.; Sugito, H.; Hasegawa, K.; Ohtsuka, K.; Fukuda, A.
Journal of Applied Polymer Science, 79, 555 (2001).
[16]
Kiskan, B.; Colak, D.; Muftuoglu, A. E.; Cianga, I.; Yagci, Y. Macromolecular
Rapid Communications, 26, 819 (2005).
[17]
Takeichi, T.; Kano, T.; Agag, T. Polymer, 46, 12172 (2005).
[18]
Chernykh, A.; Liu, J. P.; Ishida, H. Polymer, 47, 7664 (2006).
[19]
Agag, T.; Takeichi, T. Journal of Polymer Science Part a-Polymer Chemistry, 45,
1878 (2007).
[20]
Ergin, M.; Kiskan, B.; Gacal, B.; Yagci, Y. Macromolecules, 40, 4724 (2007).
[21]
Liu, Y. L.; Lin, G. C.; Wu, C. S. Journal of Polymer Science Part a-Polymer
Chemistry, 45, 949 (2007).
[22]
Chou, C. I.; Liu, Y. L. Journal of Polymer Science Part a-Polymer Chemistry, 46,
6509 (2008).
[23]
Gong, W.; Zeng, K.; Wang, L.; Zheng, S. X. Polymer, 49, 3318 (2008).
[24]
Kiskan, B.; Yagci, Y.; Ishida, H. Journal of Polymer Science Part a-Polymer
Chemistry, 46, 414 (2008).
[25]
Chernykh, A.; Agag, T.; Ishida, H. Polymer, 50, 382 (2009).
[26]
Chernykh, A.; Agag, T.; Ishida, H. Macromolecules, 42, 5121 (2009).
[27]
Ishida, H.; Rodriguez, Y. Polymer, 36, 3151 (1995).
[28]
Ishida, H.; Sanders, D. P. Macromolecules, 33, 8149 (2000).
[29]
Ishida, H.; Sanders, D. P. Polymer, 42, 3115 (2001).
[30]
Pei, D. F.; Cheng, D. R.; Gu, Y.; Cai, X. X. Thermosetting Resin, 12, 39 (1998).
[31]
Yu, D. S.; Chen, H.; Shi, Z. X.; Xu, R. W. Polymer, 43, 3163 (2002).
195
[32]
Su, Y. C.; Yei, D. R.; Chang, F. C. Journal of Applied Polymer Science, 95, 730
(2005).
[33]
Jubsilp, C.; Damrongsakkul, S.; Takeichi, T.; Rimdusit, S. Thermochimica Acta,
447, 131 (2006).
[34]
Ran, Q. C.; Li, P. Y.; Zhang, C.; Gu, Y. Journal of Macromolecular Science Part
a-Pure and Applied Chemistry, 46, 674 (2009).
[35]
Agag, T; Jin, L.; Ishida, H. Polymer, 50, 5940 (2009).
[36]
Dunkers, J.; Ishida, H. Journal of Polymer Science Part a-Polymer Chemistry, 37,
1913 (1999).
[37]
Chutayothin, P.; Ishida, H. Macromolecules, 43, 4562 (2010).
[38]
Dunkers, J.; Ishida, H. Spectrochimica Acta Part a-Molecular and Biomolecular
Spectroscopy, 51, 855 (1995).
[39]
Kumar, K. S. S.; Nair, C. P. R.; Ninan, K. N. Thermochimica Acta, 441, 150
(2006).
196
Figure A.1: Ring-Opening mechanism of P-ddm.
197
8
7
6
5
4
3
2
Chemical shift (ppm)
Figure A.2: 1H NMR spectra of P-ddm.
198
108
107
106
G', G'' (Pa)
105
104
103
102
1400C:
1600C:
1800C:
2000C:
2200C:
1
10
100
10-1 -2
10
10-1
100
G'
G'
G'
G'
G'
G''
G''
G''
G''
G''
101
Time (hours)
Figure A.3: Influence of polymerization temperature on polymerization kinetics for P-ddm.
199
a
102
tan δ
101
100
0
1
2
3
4
5
Time (hours)
b
102
tan δ
101
100
10-1
10-2
0.0
0.5
1.0
1.5
2.0
0.6
0.8
Time (hours)
c
101
tan δ
100
10-1
10-2
0.0
0.2
0.4
Time (hours)
200
100
tan δ
d
0.0
0.1
0.2
0.3
Time (hours)
100
tan δ
e
0.0
0.1
Time (hours)
Figure A.4: Polymerization kinetics of P-ddm as a function of tan δ; (a) 140oC, (b) 160oC, (c)
180oC, (d) 200oC, (e) 220oC.
201
5
Time (hours)
4
3
2
1
0
140
160
180
200
220
o
Temperature ( C)
Figure A.5: Time for P-ddm to reach the final plateau in dynamic moduli.
202
Figure A.6:
Intermolecular hydrogen-bonding structure; (a) intermolecular H-bonding
structures between monomer and opened benzoxazine, (b) opened Mannich base H-bonds to each
other.
203
107
600
6
10
500
105
104
G', G'' (Pa)
102
300
101
G'
G''
tan δ
100
10-1
10-2
tan δ
400
103
200
100
-3
10
10-4
0
-5
10
140
160
180
200
220
0
Temperature ( C)
Figure A.7: Polymerization kinetics of P-ddm as a function of temperature.
204
101
Monomer
1st tan δ peak
2nd tan δ peak
Final product
0
10
Normalized G(t) (Pa)
10-1
10-2
10-3
10-4
10-5
10-6
10-2
10-1
100
101
102
Time (s)
Figure A.8: Stress relaxation at different stages of the polymerization process of P-ddm.
205
Heat Flow (mW/g)
∆H= 311.1J/g
0
50
100
150
200
250
300
0
Temperature ( C)
Figure A.9: Non-isothermal DSC of P-ddm.
206
Heat Flow (W/g)
∆H=32.7J/g
2000C
∆H=117.5J/g
1800C
1600C
∆H=212.2J/g
1400C
∆H=244.3J/g
50
75
100
125
150
175
200
225
250
275
300
0
Temperature ( C)
Figure A.10: Non-isothermal DSC of P-ddm after crosslinking.
207
1496
1231
947
1030
Absorbance
d
c
b
a
1800
1600
1400
1200
1000
800
600
-1
Wavelength (cm )
Figure A.11: IR spectra of P-ddm with different polymerization stages at 140oC; (a) Monomer,
(b) 1st tan δ peak, (c) 2nd tan δ peak, (d) cured P-ddm.
208
1496
1478
947
Absorbance
e
d
c
b
a
1800
1600
1400
1200
1000
800
600
-1
Wavelength (cm )
Figure A.12: IR spectra of P-ddm at different polymerization temperatures and temperature
sweep; (a) 140oC, (b) 140oC, (c) 140oC, (d) 140oC, (e) T sweep.
209
BIBLIOGRAPHY
Chapter 1:
1. Shi, W., Lynd, N. A., Montarnal, D., Luo, Y., Fredrickson, G. H., Kramer, E. J., Ntaras,
C., Avgeropoulos, A., Hexemer, A., Macromolecules, 47, 2037 (2014)
2. Liu, H., Chen, F., Liu, B., Estep, G., Zhang, J., Macromolecules, 43, 6058 (2010)
3. Moungthai, S., Mahadevapuram, N., Ruchhoeft, P., Stein, G. E., ACS Appl. Mater. &
Interfaces, 4, 4015 (2012)
4. Alig, I., Pötschke, P., Lellinger, D., Skipa, T., Pegel, S., Kasaliwal, G. R., Villmow, T.,
Polymer, 53, 4 (2012)
5. Yoo, Y., Tiwari, R. R., Yoo, Y-T., Paul, D. R., Polymer, 51, 4907 (2010)
6. Andrade, R. J., Huang, R., Herbert, M. M., Chiaretti, D., Ishida, H., Schiraldi, D. A.,
Maia, J. M., Polymer, 55, 860 (2014)
7. Litchfield, D. W., Baird, D. G., Polymer, 49, 5027 (2008)
8. Djalali, R., Samson, J., Matsui, H., J. Am. Chem. Soc., 126, 7935 (2004)
9. Charleux, B., Delaittre, G., Rieger, J., Agosto, F. D., Macromolecules, 45, 6753 (2012)
10. Kozlovskaya, V., Ok, S., Sousa, A., Libera, M., Sukhishvili, S. A., Macromolecules,
36, 8590 (2003)
11. Seo, J., Lutkenhaus, J. L., Kim, J., Hammond, P. T., Char, K., Macromolecules, 40,
4028 (2007)
12. Tollar, J. E., “Interfacial surface generator”, US patent (1966)
13. Jarus, D., Hiltner, A., Baer, E., Polym. Eng. Sci., 41, 2162 (2001)
14. Ponting, M., Hiltner, A., Baer, E., Macromol. Symp., 294, 19 (2010)
210
15. Wang, H., Keum, J. K., Hiltner, A., Baer, E., Freeman, B., Rozanski, A., Galeski, A.,
Science, 323, 757 (2009)
16. Lai, C.-Y.; Hiltner, A.; Baer, E.; Korley, L. T. J. ACS Appl. Mater. & Interfaces, 4,
2218 (2012)
17. Burt, T. M.; Keum, J.; Hiltner, A.; Baer, E.; Korley, L. T. J. ACS Appl. Mater.
Interfaces. 3, 4804 (2011)
18. Cheng, W.; Gomopoulos, N.; Fytas, G.; Gorishnyy, T.; Walish, J.; Thomas, E. L.;
Hiltner, A.; Baer, E., Nano Lett., 8, 1423 (2008)
19. Tseng, J.-K.; Tang, S.; Zhou, Z.; Mackey, M.; Carr, J. M.; Mu, R.; Flandin, L.;
Schuele, D. E.; Baer, E.; Zhu, L., Polymer, 55, 8 (2014)
20. Lai, C.-Y.; Ponting, M. T.; Baer, E., Polymer, 53, 1393 (2012)
21. Singer, K. D.; Kazmierczak, T.; Lott, J.; Song, H.; Wu, Y.; Andrews, J.; Baer, E.;
Hiltner, A.; Weder, C.; OPT. EXPRESS, 16, 10358 (2008)
22. Ryan, C.; Christenson, C. W.; Valle, B.; Saini, A.; Lott, J.; Johnson, J.; Schiraldi, D.;
Weder, C.; Baer, E.; Singer, K. D.; Shan, J. Adv. Mater., 24, 5222 (2012)
23. Lott, J.; Ryan, C.; Valle, B.; Johnson, J. R.; Schiraldi, D. A.; Shan, J.; Singer, K. D.;
Weder, C. Adv. Mater., 23, 2425 (2011)
24. Zhang, G.; Lee, P. C.; Jenkins, S.; Dooley, J.; Baer, E. Polymer, 55, 663 (2014)
25. Anderson, P. D.; Dooley, J.; Meijer, H. E. H. Appl. Rheol. 16, 198 (2006)
26. Dooley, J.: Viscoelastic Flow Effects in Multilayer Polymer Coextrusion. Ph.D thesis,
Eindhoven University of Technology, Netherlands (2002)
27. Torres, A., Hrymak, A. N., Hilton, T., Rheol. Acta, 525, 513 (1993)
211
28. Hatzikiriakos, S. G., Migler, K. B. Polymer Processing Instabilities: Control and
Understanding, Marcel Dekker, New York, (2005)
29. Nazarenko, S., Snyder, J., Ebeling, T., Schuman, T., Hiltner, A., Baer, E. SPEANTEC ‘96 Proceedings, Indianapolis, Indiana, May 5-9, (1996)
30. Borzacchiello, D., Leriche, E., Blottiere, B., Guillet, J., J. Rheol. 58, 493 (2014)
31. Debbaut, B., Avalosse, T., Dooley, J., Hughes, K., J. Non-Newtonian Fluid Mech., 69,
255 (1997)
32. Yue, P., Zhou, C., Dooley, J., Feng, J. J., J. Rheol., 52, 1027 (2008)
33. Lee, P., Dooley, J., J. Elastom. Plast. 0, 1 (2013)
34. Mahdaoui, O., Agassant, J-F., Laure, P., Valette, R., Silva, L., AIP Conference
Proceedings 907, 873 (2007)
35. Huntington, B. A., Chabert, E., Rahal, Patz, J., S., Silva, J., Harris, P. J., Maia, J.,
Bonnecaze, R. T., Int. Polym. Proc., 3, 2741 (2013)
36. Joseph, D. D., J. Non-Newtonian Fluid Mech., 70, 187 (1997)
37. Harris, P. J., Patz, J., Huntington, B. A., Bonnecaze, R. T., Meltzer, D., Maia, J.,
Polym. Eng. Sci., 54, 636 (2014)
38. Li, Z., Garza, P. A. G., Baer, E., Ellison, C. J., Polymer, 53, 3245 (2012)
39. Ranade, A. P., Hiltner, A., Baer, E., J. Cell. Plast., 40, 297, (2004)
40. Dadbin, S., Noferesti, M., Frounchi, M., Macromol. Symp., 274, 22 (2008)
41. Gupta, M., Lin, Y., Deans, T., Crosby, A., Baer, E., Hiltner, A., Schiraldi, D. A.,
Polymer, 50, 598 (2009)
42. Ji, S., Yin, K., Mackey, M., Brister, A., Ponting, M., Baer, E., Opt Eng., 52, 112105
(2013)
43. Carr, J. M., Langhe, D. S., Ponting, M., Hiltner, A., Baer, E., J. Mater. Res., 27, 1326
212
(2012)
Chapter 2:
1. Wang, H., Keum, J. K., Hiltner, A., Baer, E., Freeman, B., Rozanski, A., Galeski, A.,
Science, 323, 757 (2009)
2. Mackey, M., Schuele, D. E., Zhu, L., Flandin, L., Wolak, M. A., Shirk, J. S., Hiltner,
A., Baer, E., Macromolecules, 45, 1954 (2012)
3. Lai, C.-Y., Ponting, M. T., Baer, E., Polymer, 53, 1393 (2012)
4. Carr, J. M., Mackey, M., Flandin, L., Hiltner, A., Baer, E., Polymer, 54, 1679 (2013)
5. Song, H., Singer, K., Lott, J., Wu, Y., Zhou, J., Andrews, J., Hiltner, A., Weder, C., J.
Mater. Chem., 19, 7520 (2009)
6. Jin, Y., Tai, H., Hiltner, A., Baer, E., Shirk, J. S., J. Appl. Polym. Sci., 103, 1834 (2007)
7. Ryan, C., Christenson, C. W., Valle, B., Saini, A., Lott, J., Johnson, J., Schiraldi, D.,
Weder, C., Baer, E., Singer, K., Shan, J., Adv. Mater., 24, 5222 (2012)
8. Cheng, W., Gomopoulos, N., Fytas, G., Gorishnyy, T., Walish, J., Thomas, E. L.,
Hiltner, A., Baer, E., Nano Lett., 8, 1423 (2008)
9. Dooley, J.: Viscoelastic Flow Effects in Multilayer Polymer Coextrusion. Ph.D thesis,
Eindhoven University of Technology, Netherlands (2002)
10. Anderson, P. D., Dooley, J., Meijer, H. E. H., Appl. Rheol., 16, 198 (2006)
11. Torres, A., Hrymak, A. N., Hilton, T., Rheol. Acta, 525, 513 (1993)
12. Hatzikiriakos, S. G., Migler, K. B. Polymer Processing Instabilities: Control and
Understanding, Marcel Dekker, New York, (2005)
213
13. Nazarenko, S., Snyder, J., Ebeling, T., Schuman, T., Hiltner, A., Baer, E. SPEANTEC ‘96 Proceedings, Indianapolis, Indiana, May 5-9, (1996)
14. Huang, R., Maia, J., Cox, M., U.S. Provisional Patent Appl. No. 61/901,482 (2013)
15. Sentmanat, M. L., Rheol. Acta, 43, 657 (2004)
16. Barroso, V. C., Covas, J. a., Maia, J. M., Rheol. Acta, 41, 154 (2002)
17. Harris, P. J., Patz, J., Huntington, B. A., Bonnecaze, R. T., Meltzer, D., Maia, J.,
Polym. Eng. Sci., 54, 636 (2014)
18. Thien, N. P., Tanner, R. I., J. Non-Newtonian Fluid mech., 2, 353 (1977)
19. Huntington, B. A., Chabert, E., Rahal, Patz, J., S., Silva, J., Harris, P. J., Maia, J.,
Bonnecaze, R. T., Int. Polym. Proc., 3, 2741 (2013)
20. Joseph, D. D., J. Non-Newtonian Fluid Mech., 70, 187 (1997)
21. Treffler, B., Plastics, Rubber and Composites, 34, 143 (2005)
Chapter 3:
1. Cheng, W.; Gomopoulos, N.; Fytas, G.; Gorishnyy, T.; Walish, J.; Thomas, E. L.;
Hiltner, A.; Baer, E., Nano Lett., 8, 1423 (2008)
2. Tseng, J.-K.; Tang, S.; Zhou, Z.; Mackey, M.; Carr, J. M.; Mu, R.; Flandin, L.; Schuele,
D. E.; Baer, E.; Zhu, L., Polymer, 55, 8 (2014)
3. Lai, C.-Y.; Ponting, M. T.; Baer, E., Polymer, 53, 1393 (2012)
4. Singer, K. D.; Kazmierczak, T.; Lott, J.; Song, H.; Wu, Y.; Andrews, J.; Baer, E.;
Hiltner, A.; Weder, C.; OPT. EXPRESS, 16, 10358 (2008)
5. Ryan, C.; Christenson, C. W.; Valle, B.; Saini, A.; Lott, J.; Johnson, J.; Schiraldi, D.;
Weder, C.; Baer, E.; Singer, K. D.; Shan, J. Adv. Mater., 24, 5222 (2012)
214
6. Lott, J.; Ryan, C.; Valle, B.; Johnson, J. R.; Schiraldi, D. A.; Shan, J.; Singer, K. D.;
Weder, C. Adv. Mater., 23, 2425 (2011)
7. Zhang, G.; Lee, P. C.; Jenkins, S.; Dooley, J.; Baer, E. Polymer, 55, 663 (2014)
8. Ponting, M., Hiltner, A., Baer, E., Macromol. Symp., 294, 19 (2010)
9. Wang, H., Keum, J. K., Hiltner, A., Baer, E., Freeman, B., Rozanski, A., Galeski, A.,
Science, 323, 757 (2009)
10. Dooley, J.: Viscoelastic Flow Effects in Multilayer Polymer Coextrusion. Ph.D thesis,
Eindhoven University of Technology, Netherlands (2002)
11. Yue, P., Zhou, C., Dooley, J., Feng, J. J., J. Rheol., 52, 1027 (2008)
12. Huntington, B. A., Chabert, E., Rahal, Patz, J., S., Silva, J., Harris, P. J., Maia, J.,
Bonnecaze, R. T., Int. Polym. Proc., 3, 2741 (2013)
13. Anderson, P. D., Dooley, J., Meijer, H. E. H., Appl. Rheol., 16, 198 (2006)
14. Harris, P. J., Patz, J., Huntington, B. A., Bonnecaze, R. T., Meltzer, D., Maia, J.,
Polym. Eng. Sci., 54, 636 (2014)
15. Ji, S., Yin, K., Mackey, M., Brister, A., Ponting, M., Baer, E., Opt Eng., 52, 112105
(2013).
16. Carr, J. M., Langhe, D. S., Ponting, M., Hiltner, A., Baer, E., J. Mater. Res., 27, 1326
(2012).
17. Barroso, V. C., Covas, J. a., Maia, J. M., Rheol. Acta, 41, 154 (2002)
18. Thien, N. P., Tanner, R. I., J. Non-Newtonian Fluid mech., 2, 353 (1977)
19. Silva, J., Maia, J. M., Huang, R., Meltzer, D., Cox, M., Andrade, R., Rheol Acta, 51,
947 (2012).
215
Chapter 4:
1. Furukawa, M.; Mitsui, Y.; Fukumaru, T.; Kojio, K. Polymer, 46, 10817, (2005)
2. Kojio, K.; Kugumiya, S.; Uchiba, Y.; Nishino, Y.; Furukawa, M. Polym. J., 41, 118
(2009)
3. Tang, D.; Macosko, C. W.; Hillmyer, M. A. Polym. Chem., 5, 3231 (2014)
4. Chattopadhyay, D. K.; Raju, K. V. S. N. Prog. Polym. Sci., 32, 352 (2007)
5. Chen, T. K.; Shieh, T. S.; Chui, J. Y. Macromolecules., 31, 1312, (1998)
6. Bajsic, E. G.; Rek, V.; Sendijarevic, A.; Sendijarevic, V.; Frisch K. C. J. Elastom.
Plast., 32, 162 (2000)
7. Wang, Y.; Gupta, M.; Schiraldi, D. A. J. Polym. Sci. Part B: Polym. Phys., 50, 681,
(2012).
8. Gisselfalt, K.; Helgee, B. Macromol. Mater. Eng., 288, 265, (2003)
9. Mishra, A.; Maiti, P. J. Appl. Polym. Sci., 120, 3546 (2011)
10. Prisacariu, C. Polyurethane Elastomers : From Morphology to Mechanical Aspects.
Springer: New York, (2011)
11. Tseng, J.-K.; Tang, S.; Zhou, Z.; Mackey, M.; Carr, J. M.; Mu, R.; Flandin, L.;
Schuele, D. E.; Baer, E.; Zhu, L. Polymer, 55, 8 (2014)
12. Lin, Y.; Hiltner, A.; Baer, E. Polymer, 51, 5807 (2010)
13. Zhang, G.; Lee, P. C.; Jenkins, S.; Dooley, J.; Baer, E. Polymer, 55, 663 (2014)
14. Cheng, W.; Gomopoulos, N.; Fytas, G.; Gorishnyy, T.; Walish, J.; Thomas, E. L.;
Hiltner, A.; Baer, E. Nano Lett., 8, 1423 (2008)
15. Ji S, Yin K, Mackey M, Brister A, Ponting M, Baer E. Opt Eng., 52, 112105 (2013)
16. Lai, C.-Y.; Ponting, M. T.; Baer, E. Polymer, 53, 1393 (2012)
216
17. Singer, K. D.; Kazmierczak, T.; Lott, J.; Song, H.; Wu, Y.; Andrews, J.; Baer, E.;
Hiltner, A.; Weder, C. Opt. Express., 16, 10358 (2008)
18. Ryan, C.; Christenson, C. W.; Valle, B.; Saini, A.; Lott, J.; Johnson, J.; Schiraldi, D.;
Weder, C.; Baer, E.; Singer, K. D.; Shan, J. Adv. Mater., 24, 5222 (2012)
19. Lott, J.; Ryan, C.; Valle, B.; Johnson, J. R.; Schiraldi, D. A.; Shan, J.; Singer, K. D.;
Weder, C. Adv. Mater., 23, 2425 (2011)
20. Wang, H.; Keum. J. K.; Hiltner A.; Baer, E. Macromolecules., 42, 7055 (2009)
21. Zhang, G.; Baer, E.; Hiltner, A. Polymer, 54, 4298 (2013)
22. Langhe, D. S.; Hiltner, A.; Baer, E. Polymer, 52, 5879 (2011)
23. Wang, H.; Keum, J. K.; Hiltner, A.; Baer, E.; Freeman, B.; Rozanski, A.; Galeski, A.
Science, 323, 757 (2009)
24. Carr, J. M.; Mackey, M.; Flandin, L.; Schuele, D.; Zhu, L.; Baer, E. J. Polym. Sci.
Part B: Polym. Phys., 51, 882 (2013)
25. Burt, T. M.; Keum, J.; Hiltner, A.; Baer, E.; Korley, L. T. J. ACS Appl. Mater.
Interfaces., 3, 4804 (2011)
26. Sheng, D.; Tan, J.; Liu, X.; Wang, P.; Yang, Y. J. Mater. Sci., 46, 6508 (2011)
27. Lai, C.-Y.; Hiltner, A.; Baer, E.; Korley, L. T. J. ACS Appl. Mater. & Interfaces, 4,
2218 (2012)
28. Carr, J. M.; Langhe, D. S.; Ponting, M. T.; Hiltner, A.; Baer, E. J. Mater. Res., 27,
1326 (2012)
29. Carr, J. M.; Mackey, M.; Flandin, L.; Hiltner, A.; Baer, E. Polymer, 54, 1679 (2013)
30. Koerner, H.; Kelley, J. J.; Vaia, R. A. Macromolecules, 41, 4709 (2008)
31. Shibayama, M.; Inoue, M.; Yamamoto, T.; Nomura,S. Polymer., 31, 749 (1990)
217
32. Huang, R.; Patz, J.; Silva, J.; Maia, J. M.; Huntington, B. A.; Bonnecaze, R. T.; Cox,
M. ANTEC 2014 Proceedings. April 27-30, Las Vegas, accepted
33. Huang, R.; Patz, J.; Silva, J.; Andrade, R.; Harris, P.; Yin, K.; Huntington, B.;
Bonnecaze, R.; Cox, M.; Maia, J. M. Int. Polym. Proc. submitted
34. Barroso, V.; Covas, J. A.; Maia, J. M. Rheol. Acta, 41, 154 (2002)
35. Harris, P. J.; Patz, J.; Huntington, B. A.; Bonnecaze, R. T.; Meltzer, D.; Maia, J.
Polym. Eng. & Sci., 54, 636 (2014)
36. Silva, J.; Andrade, R.; Huang, R.; Liu, J.; Meltzer, D.; Cox, M.; Maia, J. M. J nonNewtonian Fluid Mech. 2014, submitted.
37. Silva, J.; Maia, J. M.; Huang, R.; Meltzer, D.; Cox, M.; Andrade, R. Rheol. Acta., 51,
947 (2012)
38. Sakurai, S.; Yoshida, H.; Hashimoto, F.; Shibaya, M.; Ishihara, H.; Yoshihara, N.;
Nishitsuji, S.; Takenaka, M. Polymer, 50, 1566 (2009)
39. Coleman, M.M.; Lee, K. H.; Skrovanek, D. J.; Painter, P. C. Macromolecules., 19,
2149 (1986)
40. Ponting, M.; Burt, T. M.; Korley, L. T. J.; Andrews, J.; Hiltner, A.; Baer, E. Ind. &
Eng. Chem. Res., 49, 12111 (2010)
Chapter 5:
1. Chen, Y., Kushner, A. M., Williams, G. A., Guan, Z., Nat Chem, 4, 467 (2012)
2. Shi, W., Lynd, N. A., Montarnal, D., Luo, Y., Fredrickson, G. H., Kramer, E. J.,
Macromolecules, 47,2037 (2014)
3. George, J., Varughese, K. T., Thomas, S., Polymer, 41, 1507 (2000)
218
4. Sonnenschein, M. F., Ginzburg, V. V., Schiller, K. S., Wendt, B. L., Polymer, 54, 1350
(2013)
5. Thomas, S., George, A., Eur. Polym. J., 28, 1451 (1992)
6. George, S., Ramamurthy, K., Anand, J., Groeninckx, G., Varughese, K.T., Thomas, S.,
Polymer, 40, 4325 (1999)
7. Kollosche, M., Kofod, G., Appl. Phys. Lett., 96, 071904 (2010)
8. Armstrong, S., Freeman, B., Hiltner, A., Baer, E., Polymer, 53, 1383 (2012)
9. Khan, U., May, P., O’Neill, A., Vilatela, J. J., Windle, A. H., Coleman, J. N., Small, 7,
1579 (2011)
10. Lin, Z., Chen, C., Guan, Z., Tan, S., Zhang, X., Al, L. J., J. Appl. Polym. Sci., 122,
2789 (2011)
11. Panaitescu, D. M., Vuluga, Z., Radovici, C., Nicolae, C., Polym. Test., 31, 355 (2012)
12. Vuluga, Z., Panaitescu, D. M., Radovici, C., Nicolae, C., Iorga, M. D., Polym. Bull.,
69, 1073 (2012)
13. Gupta, A. K., Purwar, S. N. J., J. Appl. Polym. Sci., 29, 1595 (1984)
14. Gupta, A. K., Purwar, S. N. J., J. Appl. Polym. Sci., 29, 3513 (1984)
15. Gupta, A. K., Purwar, S. N. J., J. Appl. Polym. Sci., 31, 535 (1986)
16. Gupta, A. K., Srinivasan, K. R. J., J. Appl. Polym. Sci., 47, 167 (1993)
17. Wilkinson, A. N., Laugel, L., Clemens, M. L., Harding, V. M., Marin, M., Polymer,
40, 4971 (1999)
18. Kusmono, Ishak, Z. A. M., Chow, W. S., Takeichi, T., Rochmad., Eur. Polym. J., 44,
1023 (2008)
19. Stribeck, N., Fakirov, S., Macromolecules, 34, 7758 (2001)
219
20. Elleuch, R., Elleuch, K., Salah, B., Zahouani, H., Mater. Des., 28, 824 (2007)
21. Yu, Z., Lei, M., Ou, Y.; Yang, G., Polymer, 43, 6993 (2002)
22. Saikrasun, S., Amornsakchai, T., Sirisinha, C., Meesiri, W., Bualek-Limcharoen, S.,
Polymer, 40, 6437 (1999)
23. Li, Y., Oono, Y., Kadowaki, Y., Inoue, T., Nakayama, K., Shimizu, H.,
Macromolecules, 39, 4195 (2006)
24. Chang, K., Robertson, M. L., Hillmyer, M. A., ACS Appl. Mater. Interfaces, 1, 2390
(2009)
25. Ohlsson, B., Tornell, B., Polym. Eng. Sci., 36, 1547 (1996)
26. Sengers, W. G. F., Sengupta, P., Noordermeer, J. W. M., Picken, S. J., Gotsis, A. D.,
Polymer, 45, 8881 (2004)
27. Jha, A., Dutta, B., Bhowmick, A. K. J., J. Appl. Polym. Sci., 74, 1490 (1999)
28. Sengers, W. G. F., Wübbenhorst, M., Picken, S. J., Gotsis, A. D., Polymer. 46, 6391
(2005)
29. Sengupta, P., Noordermeer, J. W. M., Polymer, 46, 12298 (2005)
30. Sengupta, P., Noordermeer, J. W. M., Macromol. Rapid Commun., 26, 542 (2005)
Appendix:
1. Ishida H. in "Handbook of Benzoxazine Resins," Ishida H and Agag T, Eds. Elsevier,
Amsterdam, pp. 3-81 (2011).
2. Pilato L. "Phenolic Resins: A Century of Progress." Springer, New York (2010).
3. Ishida, H.; Allen, D. J. Journal of Polymer Science Part B-Polymer Physics, 34, 1019
(1996).
220
4. Low, H. Y.; Ishida, H. Journal of Polymer Science Part B-Polymer Physics, 36, 1935
(1998).
5. Low, H. Y.; Ishida, H. Journal of Polymer Science Part B-Polymer Physics, 37, 647
(1999).
6. Kim, H. D.; Ishida, H. Journal of Applied Polymer Science, 79, 1207 (2001).
7. Wu, Y. Q.; Zeng, M.; Jin, H. Y.; Xu, Q. Y.; Lu, L. Y.; Wang, J.; Hou, S. E. Tribology
Letters, 47, 389 (2012).
8. Espinosa, M. A.; Galia, M.; Cadiz, V. Polymer, 45, 6103 (2004).
9. Ishida, H.; Low, H. Y. Macromolecules, 30, 1099 (1997).
10. Liu, X.; Gu, Y. Acta Polymerica Sinica, 5, 612 (2000).
11. Kim, H. J.; Brunovska, Z.; Ishida, H. Polymer, 40, 1815 (1999).
12. Ning, X.; Ishida, H. Journal of Polymer Science Part a-Polymer Chemistry, 32, 1121
(1994).
13. Wirasate, S.; Dhumrongvaraporn, S.; Allen, D. J.; Ishida, H. Abstracts of Papers of
the American Chemical Society, 213, 81-Pmse (1997).
14. Kim, H. D.; Ishida, H. Journal of Physical Chemistry A, 106, 3271 (2002).
15. Kimura, H.; Matsumoto, A.; Sugito, H.; Hasegawa, K.; Ohtsuka, K.; Fukuda, A.
Journal of Applied Polymer Science, 79, 555 (2001).
16. Kiskan, B.; Colak, D.; Muftuoglu, A. E.; Cianga, I.; Yagci, Y. Macromolecular Rapid
Communications, 26, 819 (2005).
17. Takeichi, T.; Kano, T.; Agag, T. Polymer, 46, 12172 (2005).
18. Chernykh, A.; Liu, J. P.; Ishida, H. Polymer, 47, 7664 (2006).
221
19. Agag, T.; Takeichi, T. Journal of Polymer Science Part a-Polymer Chemistry, 45,
1878 (2007).
20. Ergin, M.; Kiskan, B.; Gacal, B.; Yagci, Y. Macromolecules, 40, 4724 (2007).
21. Liu, Y. L.; Lin, G. C.; Wu, C. S. Journal of Polymer Science Part a-Polymer
Chemistry, 45, 949 (2007).
22. Chou, C. I.; Liu, Y. L. Journal of Polymer Science Part a-Polymer Chemistry, 46,
6509 (2008).
23. Gong, W.; Zeng, K.; Wang, L.; Zheng, S. X. Polymer, 49, 3318 (2008).
24. Kiskan, B.; Yagci, Y.; Ishida, H. Journal of Polymer Science Part a-Polymer
Chemistry, 46, 414 (2008).
25. Chernykh, A.; Agag, T.; Ishida, H. Polymer, 50, 382 (2009).
26. Chernykh, A.; Agag, T.; Ishida, H. Macromolecules, 42, 5121 (2009).
27. Ishida, H.; Rodriguez, Y. Polymer, 36, 3151 (1995).
28. Ishida, H.; Sanders, D. P. Macromolecules, 33, 8149 (2000).
29. Ishida, H.; Sanders, D. P. Polymer, 42, 3115 (2001).
30. Pei, D. F.; Cheng, D. R.; Gu, Y.; Cai, X. X. Thermosetting Resin, 12, 39 (1998).
31. Yu, D. S.; Chen, H.; Shi, Z. X.; Xu, R. W. Polymer, 43, 3163 (2002).
32. Su, Y. C.; Yei, D. R.; Chang, F. C. Journal of Applied Polymer Science, 95, 730
(2005).
33. Jubsilp, C.; Damrongsakkul, S.; Takeichi, T.; Rimdusit, S. Thermochimica Acta, 447,
131 (2006).
34. Ran, Q. C.; Li, P. Y.; Zhang, C.; Gu, Y. Journal of Macromolecular Science Part aPure and Applied Chemistry, 46, 674 (2009).
222
35. Agag, T; Jin, L.; Ishida, H. Polymer, 50, 5940 (2009).
36. Dunkers, J.; Ishida, H. Journal of Polymer Science Part a-Polymer Chemistry, 37,
1913 (1999).
37. Chutayothin, P.; Ishida, H. Macromolecules, 43, 4562 (2010).
38. Dunkers, J.; Ishida, H. Spectrochimica Acta Part a-Molecular and Biomolecular
Spectroscopy, 51, 855 (1995).
39. Kumar, K. S. S.; Nair, C. P. R.; Ninan, K. N. Thermochimica Acta, 441, 150 (2006).
223