PLACE IN RETURN BOX to remove this checkout from your record.

TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

 

 

DAIEDUE

DAIEDUE

DAIEDUE

 

M}

‘1 l) ‘1 A

 

‘7 “7 111,31”

 

 

 

 

 

 

 

 

 

 

 

 

 

5108 K:lProjIAcc&Pres/CIRCIDaIeDue.indd

 

DEVELOPMENT OF POLYLACTIC ACID-BASED MATERIALS THROUGH
REACTIVE MODIFICATION

By

Alison Camille Fowlks

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Materials Science and Engineering

2009

ABSTRACT

DEVELOPMENT OF POLYLACTIC ACID-BASED MATERIALS THROUGH
REACTIVE MODIFICATION

BY

Alison Camille Fowlks

Polylactic acid (PLA)-based systems have shown to be of great potential
for the development of materials requiring biobased content, biodegradation, and
sufﬁcient properties. The efforts in this study are directed toward addressing the
current research need to overcome some of the inherent drawbacks of PLA. To
meet this need, reactive extrusion was employed to develop new materials based
on PLA by grafting, compounding, and polymer blending.

In the ﬁrst part of this work, maleic anhydride (MA) was grafted onto PLA
by reactive extrusion. Two structurally different peroxides were used to initiate
grafting and results were reported on the basis of grafting, molecular weight, and
thermal behavior. An inverse relationship between degree of grafting and
molecular weight was established. It was also found that, regardless of peroxide
type, there is an optimum peroxid-to-MA ratio of 0.5:2 that promotes maximum
grafting, beyond which degradation reactions become predominant. Overall, it
was found that the maleated copolymer (MAPLA) could be used as an interfacial
modiﬁer in FLA-based composites. Therefore, MAPLA was incorporated into
PLA-talc composites in varying concentrations. The inﬂuence of the MAPLA
addition on the mechanical and thermal behavior was investigated. When added

in an optimum concentration, MAPLA improved the tensile strength and

crystallization of the composite. Furthermore, microscopic observation conﬁrmed
the compatibilization effect of MAPLA in PLA-talc composites.

Vinyltrimethoxysilane was free-radically grafted onto the backbone of PLA
and subsequently moisture crosslinked. The effects of monomer, initiator, and
catalyst concentration on the degree of crosslinking and the mechanical and
thermal properties were investigated. The presence of a small amount of catalyst
showed to be a major contributor to the crosslinking formation in the time frame
investigated, shown by an increase in gel content and decrease in crystallinity.
Furthermore, thennogravimetric analysis helped to establish a correlation
between crosslinking density and thermal decomposition—providing conclusive
evidence of improved thermal stability as a result of the crosslinking reaction.
The mechanical properties indicated that the crosslinked PLA was developed
without embrittlement or a reduction in tensile strength.

In the ﬁnal section of this work a blend comprising of PLA and PBAT was
reactively compatibilized via an in situ transesteriﬁcation reaction and blown ﬁlms
were produced. The compatibilized ﬁlms demonstrated enhanced properties
compared to those of the physical blends. The most signiﬁcant improvements
were observed in the composition where PBAT was the majority phase, PLA was
the minor phase, and catalyst was added in low concentration. Morphological
observation conﬁrmed interaction between polymer phases by improved
dispersion and signiﬁcant reduction in domain size which inferred the forniation

of an interfacial copolymer.

Dedicated to

My Parents

iv

ACKNOWLEDGEMENTS

First, I would like to thank my advisor, Dr. Ramani Narayan, for his
mentoring, guidance, and support during the course of pursuing my Ph.D His
creative ideas and invaluable insights have been instrumental in my growth as an
engineer and independent researcher. Next, I acknowledge my dissertation
committee: Dr. Rafael Auras, Dr. Martin Crimp, and Dr. Bruce Dale for the
interest, support, and constructive advice given in my project and for serving on
my Ph.D. guidance committee. I would like to extend a special thank you to Ken
Farminer and Dr. Dan Graiver for their encouragement and useful suggestions.
To those who have assisted in speciﬁc areas pertaining to my research project, I
extend my heartfelt gratitude: Dr. Per Askeland for his FT IR and XPS expertise;
and Dr. Ewa Danielewicz and Dr. Alicia Pastor for their counsel and assistance in
electron microscopy. I would like to recognize my colleagues in the Biobased
Materials Research Group. Their motivation and assistance during these last four
and a half years is gratefully acknowledged. My family and friends deserve
special recognition for all of their prayers, support, and never-ending love. This
work would not have been possible without the love and support of my parents.
My father, Edison, who is one of the wisest men I know, piqued my interest in
science at an early age; and my mother, Corinne, who has believed in me from
the very beginning and whose assistance during this endeavor has been
invaluable. Together they have always taught me to follow my dreams. My
brother, Kevin, deserves special mention for his support and encouragement and

for being the best big brother anyone could have. Last, but not least, a special

V

word of thanks goes to my ﬁancé, Michael, who on countless occasions
sacriﬁced his weekends to help me extrude, and also for his patience and

understanding over the past three years.

vi

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................ X
LIST OF FIGURES .............................................................................................. xi
CHAPTER 1 .......................................................................................................... 1
Rationale ........................................................................................................... 1
Scope of the study ............................................................................................. 6
Structure of THE Dissertation ............................................................................ 6
REFERENCES .................................................................................................. 9
CHAPTER 2 ........................................................................................................ 11
PLA Synthesis .............................................................................................. 13
PLA Properties ............................................................................................. 16
PLA Biodegradation ..................................................................................... 19
Polylactic Acid Processing ........................................................................... 21
PLA Limitations ............................................................................................ 21
REFERENCES ................................................................................................ 23
CHAPTER 3 ........................................................................................................ 26
Reactive Extrusion ........................................................................................... 26
Titrimetric Analysis .......................................................................................... 28
Gel Permeation Chromatography (GPC) ......................................................... 29
Thermal Characterization ................................................................................ 3O
Differential Scanning Calorimetry (DSC) ...................................................... 31
Thennogravimetric Analysis (TGA) .............................................................. 32
Dynamic Mechanial Analysis (DMA) ............................................................ 32
Mechanical Characterization ........................................................................... 34
Tensile Testing ............................................................................................. 34
Impact Testing ............................................................................................. 36
Melt Flow Index (MFI) ...................................................................................... 37
Gel Content and Swell Ratio ........................................................................... 38
Gelation ....................................................................................................... 38
Degree of Swelling ....................................................................................... 38
Electron Microscopy ........................................................................................ 39
Scanning Electron Microscopy (SEM) .......................................................... 39
Transmission Electron Microscopy (TEM) ................................................... 40
X-Ray Photoelectron Spectroscopy (XPS) ...................................................... 41
Fourier Transform Infrared (FTIR) Spectroscopy ............................................. 42
REFERENCES ................................................................................................ 43
CHAPTER 4 ........................................................................................................ 45
INTRODUCTION ............................................................................................. 45
EXPERIMENTAL ............................................................................................. 48
Materials ...................................................................................................... 48

vii

Methods ....................................................................................................... 49

RESULTS AND DISCUSSION ........................................................................ 53
MA Grafting Reaction ................................................................................... 53
Proof of MA Grafting by Fourier Transform Infrared Spectroscopy (FTIR)...56
Grafting Efﬁciency and Molecular Weight .................................................... 57
Effect of Initiator Concentration on the Grafting Yield of MAPLA ................. 58
Effect of Initiator Concentration on the Molecular Weight of MAPLA ........... 60
Effect of MA Concentration on the Grafting Yield of MA .............................. 62
Effect of MA Concentration on the Molecular Weight of MAPLA ................. 64
Thermal Stability of MAPLA ......................................................................... 65
Differential Scanning Calorimetry ................................................................. 69

CONCLUSIONS .............................................................................................. 72

REFERENCES ................................................................................................ 74

CHAPTER 5 ........................................................................................................ 76

INTRODUCTION ............................................................................................. 76

THEORETICAL BACKGROUND ..................................................................... 81
Rule of Mixtures ........................................................................................... 82
Adhesion Theory .......................................................................................... 83

EXPERIMENTAL ............................................................................................. 88
Materials ...................................................................................................... 88
Methods ....................................................................................................... 88

RESULTS AND DISCUSSION ........................................................................ 92
Proposed Reaction Mechanism ................................................................... 92
Differential Scanning Calorimetry ................................................................. 93
Mechanical Properties ................................................................................. 96
Therrnomechanical Properties ................................................................... 100
Scanning Electron Microscopy ................................................................... 107

CONCLUSIONS ............................................................................................ 1 17

REFERENCES .............................................................................................. 118

CHAPTER 6 ...................................................................................................... 121

INTRODUCTION ........................................................................................... 121
PLA Crosslinking Background ................................................................... 124
Silane Crosslinking .................................................................................... 127
Network Theory .......................................................................................... 129

EXPERIMENTAL ........................................................................................... 131
Materials .................................................................................................... 131
Methods ..................................................................................................... 1 32

RESULTS AND DISCUSSION ...................................................................... 137
Proposed Reaction Mechanism ................................................................. 137
Role of the Catalyst .................................................................................... 139
Effect of Silane Concentration on the Degree of Crosslinking ................... 140
Effect of Initiator Concentration on Gel Content ......................................... 141
Effect of Catalyst Concentration on Gel Content ........................................ 142
Effect of Crosslinking Time on Gel Content ............................................... 143

viii

Rate of Crosslinking ................................................................................... 147

Degree of Swelling ..................................................................................... 148
Proof of Grafting by XPS ............................................................................ 149
Effect of Crosslinking on Melt Flow Behavior ............................................. 152
Thermal Properties .................................................................................... 1 53
Thermal Stability ........................................................................................ 1 58
Mechanical Properties ............................................................................... 161
CONCLUSIONS ............................................................................................ 162
REFERENCES .............................................................................................. 164
CHAPTER 7 ...................................................................................................... 168
INTRODUCTION ........................................................................................... 168
PLA-Based Blends ..................................................................................... 169
Biodegradable PLA-Based Blends ............................................................. 171
Theoretical Background ................................................................................. 1 75
Basic Thermodynamics of Polymer Blending ............................................. 175
Miscibility Determination ............................................................................ 178
Polymer Blend Morphology ........................................................................ 179
Tailored Polymer Blend Properties ............................................................ 182
Compatibilized Polymer Blends ................................................................. 183
Reactive Compatibilization ......................................................................... 186
EXPERIMENTAL ........................................................................................... 187
Materials .................................................................................................... 187
Methods ..................................................................................................... 187
RESULTS AND DISCUSSION ...................................................................... 191
Proposed Transesteriﬁcation Reaction ...................................................... 191
Mechanical Properties of Blends ................................................................ 192
Molecular Weight of PBAT/PLA Blends ..................................................... 199
Thermal Properties .................................................................................... 202
Blend Morphology ...................................................................................... 217
CONCLUSIONS ............................................................................................ 232
REFERENCES .............................................................................................. 234
CHAPTER 8 ...................................................................................................... 238
Conclusions ................................................................................................... 238
Recommendations for future work ................................................................. 240

ix

LIST OF TABLES

Table 2.1: Mechanical Properties of PLA ........................................................... 19

Table 4.1: Molecular Weight Data and Graft Yield as a Function of Initiator
Concentration ...................................................................................................... 58

Table 4.2: Molecular Weight Data and Graft Yield as a Function of MA
Concentration ...................................................................................................... 62

Table 4.3: DSC Data of L101-initiated MAPLA with Varying L101 Concentrations
............................................................................................................................ 70

Table 4.4: DSC Data of L101-initiated MAPLA with Varying MA Concentrations71
Table 4.5: DSC Data of T301-initiated MAPLA with Varying T301 Concentrations

............................................................................................................................ 72
Table 4.6: DSC Data of T301-initiated MAPLA with Varying MA Concentrations

............................................................................................................................ 72
Table 5.1: Extrusion Conditions for Compounding PLA/Talc Composites ........... 89

Table 5.2: Injection Molding Temperatures for PLA and PLA/Talc Composites..89

Table 5.3: Melting, Crystallization, and Glass Transition Temperatures for

PLA/Talc Composites .......................................................................................... 95
Table 5.4: Tensile Properties of PLA/Talc Composites ....................................... 98
Table 5.5: Heat Distortion Temperatures of PLA/Talc Composites ................... 101
Table 6.1: VTMOS Grafting Sample Compositions ........................................... 133
Table 6.2: Extrusion Processing Conditions for VTMOS Grafting ..................... 133
Table 6.3: Injection Molding Processing Conditions for VTMOS Grafting ......... 134
Table 6.3: Crosslinking Rate Constants of Crosslinked Samples ..................... 148
Table 6.4: Melt Flow Index Values for Crosslinked PLA ................................... 153

Table 6.5: DSC Data of Crosslinked PLA with Varying Initiator Concentrations155

Table 6.6: DSC Data of Crosslinked PLA with varying VTMOS Concentrations
.......................................................................................................................... 1 55

Table 6.7: DSC Data of Crosslinked PLA with Varying Catalyst Concentrations
.......................................................................................................................... 1 57

X

Table 7.1: PBAT/PLA Blend Compositions ....................................................... 188

Table 7.2: Extrusion Processing Conditions for PBAT/PLA Blends .................. 188
Table 7.3: Blown Film Processing Conditions for PBAT/PLA Blends ................ 189
Table 7.4: Molecular Weights and Molecular Weight Distributions of PBAT/PLA

Blends ............................................................................................................... 202
Table 7.5: DSC Data of PBAT/PLA Blends ....................................................... 205
Table 7.6: Thermogravimetric Parameters of PBAT/PLA Blends ...................... 217

xi

LIST OF FIGURES

Figure 2.1: Schematic illustration of the stereochemically different lactic acid
monomers and their corresponding cyclic dimer lactide monomers .................... 13

Figure 2.2: Synthesis of low molecular weight PLA via a condensation
polymerization reaction ....................................................................................... 14

Figure 2.3: Synthesis of high molecular weight PLA via azeotropic condensation,
chain extension and ring-opening polymerization ............................................... 16

Figure 4.1: Structure of maleic anhydride (MA) ................................................... 49
Figure 4.2: Structure of 2, 5-dimethyl-2, 5-di-(tert-butylperoxy) hexane (L101)...49
Figure 4.3: Structure of 3,6,9-triethyl-3,6,9-trimethyl-1,4,7—triperoxonane (T301)

............................................................................................................................ 49
Figure 4.4: Proposed decomposition reaction of T301 ........................................ 54
Figure 4.5: Proposed reaction for grafting MA onto PLA ..................................... 55
Figure 4.6: FTIR spectra of PLA and MAPLA initiated with T301 verifying MA
grafting ................................................................................................................ 57
Figure 4.7: Effect of initiator concentration on MA grafting yield at 2% MA loading
............................................................................................................................ 60
Figure 4.8: Effect of initiator concentration on the number average molecular
weight of MAPLA at 2% MA loading ................................................................... 61
Figure 4.9: Effect of MA loading on the MA grafting yield at 0.5% initiator
concentration ...................................................................................................... 63
Figure 4.10: Effect of MA loading on the number average molecular weight at
0.5% initiator concentration ................................................................................. 65

Figure 4.11: Effect of L101 on the thermal stability of PLA prepared with 2% MA
............................................................................................................................ 66

Figure 4.12: Effect of MA concentration on the thermal stability of PLA prepared
with 0.5% L101 ................................................................................................... 67

Figure 4.13: Effect of MA concentration on the thermal stability of PLA prepared
with 0.5% T301 ................................................................................................... 68

xii

Figure 4.14: Effect of T301 concentration on the thermal stability of PLA prepared

with 2% MA ......................................................................................................... 69
Figure 5.1: Schematic diagram of a contact angle and its surface tension

components. ....................................................................................................... 85
Figure 5.2: Schematic diagram of chemical bonding between interfaces ........... 87
Figure 5.3: Schematic diagram of mechanical interlocking ................................. 87
Figure 5.4: Schematic diagram of interdiffusion adhesion mechanism ............... 88

Figure 5. 5: Proposed reaction between PLA, talc, and MAPLA yielding a

monoester on the surface of the talc particle ...................................................... 92
Figure 5.6: DSC thennograms of PLA/talc composites revealing the effect of

MAPLA on the melting behavior .......................................................................... 95
Figure 5.7: DSC thennograms of PLA/talc composites revealing the effect of

MAPLA on the crystallization temperature .......................................................... 96
Figure 5.8: Impact properties of PLA/talc composites ....................................... 100
Figure 5.9: Storage modulus curves of PLA and PLA/talc composites ............. 103
Figure 5.10: Tan 6 curves of PLA and PLA/talc composites ............................. 104
Figure 5.11: Loss modulus curves of PLA and PLA/talc composites ................ 106

Figure 5.12: Backscattered electron image of PLA-talc composite containing no
MAPLA .............................................................................................................. 1 08

Figure 5.13: Backscattered electron image of PLA-talc composite containing 3%
MAPLA .............................................................................................................. 1 09

Figure 5.14: Backscattered electron image of PLA-talc composite containing 5%
MAPLA .............................................................................................................. 1 10

Figure 5.15: Backscattered electron image of PLA-talc composite containing 10%
MAPLA .............................................................................................................. 1 1 1

Figure 5.16: Secondary electron SEM micrograph of PLA-talc composite with no
added MAPLA indicating little or no adhesion between talc particles and PLA
matrix as evidenced by platelet pull-out ............................................................ 114

Figure 5.17: Secondary electron SEM micrograph of PLA-talc composite
containing 3% MAPLA denoting slight to moderate adhesion between talc and
PLA ................................................................................................................... 1 15

xiii

Figure 5.18: Secondary electron SEM micrograph of PLA-talc composite
containing 5% MAPLA revealing enhanced adhesion and a more ductile fracture

.......................................................................................................................... 1 16
Figure 5.19: Secondary electron SEM micrograph of PLA-talc composite
containing 10% MAPLA displaying a ﬂat surface texture, indicating brittle
composite failure in the PLA matrix ................................................................... 117

Figure 6.1: Schematic illustration of a crosslinked and uncrosslinked polymer.123
Figure 6.2: Stmcture of vinyl trimethoxysilane (VTMOS) .................................. 132
Figure 6.3: Structure of 2, 5-dimethyl-2, 5-di-(tert-butylperoxy) hexane (L101).132
Figure 6.7 Expected mechanism of DBTDL reaction33 ...................................... 140

Figure 6.8: Effect of VTMOS concentration on the gel content of samples
prepared with 0.1% L101 after 28 days of crosslinking ..................................... 141

Figure 6.9: Effect of L101 concentration on the gel content of samples prepared
with 3% VTMOS after 28 days of crosslinking .................................................. 142

Figure 6.10: Effect of catalyst concentration on the gel content of samples
prepared with 3% VTMOS and 0.1% L101 after 28 days of crosslinking .......... 143

Figure 6.11: Effect of crosslinking time and VTMOS concentration on the gel
content of crosslinked samples (a) no DBTDL (b) 0.05% DBTDL ..................... 145

Figure 6.13: Linear dependence of crosslinking time on the relative gel content of
samples prepared in the presence and absence of catalyst ............................. 147

Figure 6.14: Reciprocal swelling ratio as a function of initiator concentration
determined after 28 days of crosslinking ........................................................... 149

Figure 6.15: XPS spectrum of extruded PLA. ................................................... 150
Figure 6.16: XPS spectrum of crosslinked PLA prepared with 0.5% VTMOS ...150 ,
Figure 6.17: XPS spectrum of crosslinked PLA prepared with 1% VTMOS ...... 151
Figure 6.18: XPS spectrum of crosslinked PLA prepared with 3% VTMOS ...... 151

Figure 6.19: TGA traces of crosslinked PLA with varying initiator concentrations
.......................................................................................................................... 1 59

Figure 6.20: TGA traces of crosslinked PLA with varying initiator concentrations
prepared with 0.05% catalyst. ........................................................................... 159

xiv

Figure 6.21: TGA traces of crosslinked PLA with varying catalyst concentrations
at a constant 3% VTMOS and 0.1% L101 loadings. ......................................... 160

Figure 7.1: Phase diagrams showing transition behavior in polymer blends 26 .178

Figure 7.2: Schematic representation of the effect of viscosity ratio and blend

composition on the morphology development of a polymer blend .................... 181
Figure 7.3: Property-composition relationship in polymer blends ...................... 183
Figure 7.4: Structure of poly(butylene adipate co-terephthalate) (PBAT) .......... 187
Figure 7.5: Structure of stannous 2-ethylhexanoate (SnOctz) ........................... 187
Figure 7.6: Possible transesteriﬁcation reaction between PLA and PBAT ........ 192
Figure 7.7: Tensile strength of PBAT/PLA ﬁlms tested in the machine directior'I93
Figure 7.8: Tensile strength of PBAT/PLA ﬁlms tested in the transverse directzogr:l
Figure 7.9: Elongation at break of PBAT/PLA ﬁlms tested in the machine
direction ............................................................................................................ 1 95
Figure 7.10: Elongation at break of PBAT/PLA ﬁlms tested in the transverse
direction ............................................................................................................ 195
Figure 7.11: Tear resistance of PBAT/PLA ﬁlms tested in the machine directio1n9 8
Figure 7.12: Tear resistance of PBAT/PLA ﬁlms tested in the transverse
direction ............................................................................................................ 1 98

Figure 7.13: GPC traces of PBAT/PLA blends for the following compositions: .201

(a) 30:70; (b) 70:30; (c) 50:50 ........................................................................... 201
Figure 7.14: Storage modulus curves of PBAT/PLA physical blends ................ 207
Figure 7.15: Loss modulus curves of PBAT/PLA physical blends ..................... 208
Figure 7.16: Tan delta curves of PBAT/PLA physical blends ............................ 209

Figure 7.17: Storage modulus curves of PBAT/PLA curves with varying catalyst
concentrations ................................................................................................... 21 1

Figure 7.18: Loss modulus curves of PBAT/PLA curves with varying catalyst
concentrations ................................................................................................... 21 2

XV

Figure 7.19: Tan delta curves of PBAT/PLA curves with varying catalyst
concentrations ................................................................................................... 21 3

Figure 7.20: TGA thennograms of PBAT/PLA physical blends ......................... 214

Figure 7.21: TGA thennograms of PBAT/PLA blends showing the effect of
catalyst concentration on the decomposition behavior ...................................... 215

Figure 7.22: Differential thermogravimetric curves of 70:30 PBAT/PLA ﬁlms
showing effect of catalyst concentration on the decomposition temperatures...216

Figure 7.23: TEM micrograph of the 70:30 PBAT/PLA physical blend .............. 219
Figure 7.24: TEM micrograph of 50:50 PBAT/PLA physical blend .................... 220
Figure 7.25: TEM micrograph of 30:70 PBAT/PLA physical blend .................... 221
Figure 7.26: TEM micrograph of 70:30 PBAT/PLA blend prepared with 0.05%

SnOctz ............................................................................................................... 223
Figure 7.27: TEM micrograph of 50:50 PBAT/PLA blend prepared with 0.05%

ShOCtz ............................................................................................................... 224
Figure 7.28: TEM micrograph of 30:70 PBAT/PLA blend prepared with 0.05%

SHOOT; .............................................................................................................. 225
Figure 7.29: TEM micrograph of 70:30 PBAT/PLA physical blend. ................... 227

Figure 7.30: TEM micrograph of 70:30 PBAT/PLA blend prepared with 0.05%
SnOctz ............................................................................................................... 228

Figure 7.31: TEM micrograph of 70:30 PBAT/PLA blend prepared with 0.1%
SnOctz ............................................................................................................... 229

Figure 7.32: TEM micrograph of 70:30 PBAT/PLA blend prepared with 0.5%
SnOctz ............................................................................................................... 231

Figure 7.33: TEM micrograph of 70:30 PBAT/PLA blend prepared with 0.5%
SnOctz taken at 40,000X magniﬁcation ............................................................. 232

xvi

Chapter 1

 

INTRODUCTION

 

RATIONALE

Over the last century, a vast array of synthetic polymers was introduced
and has since been incorporated into essentially all aspects of modern society.
The development of these materials, derived from petroleum resources, has
played a central role in transforming daily life,1 replacing many natural materials
in an unprecedented range of applications. Several characteristics and qualities
of plastics account for their utility and ubiquitous presence in society; among
these are their convenience, versatility, low cost, low weight, and high strength.2
Additionally, the once-thought notion of an abundant supply of petroleum-based
monomers as basicbuilding blocks contributed to their unrelenting growth.
Likewise, their durability and resistance to degradation made them all the more

attractive.3 However, this non-degradability is what prevents their disposal in safe

and ecologically sound ways.4 Although some of these materials are recyclable,
it is not always economically, nor energetically viable to do so and, as a result,
they end up in landﬁlls occupying valuable space. According to a 2006 study by
the Environmental Protection Agency,5 each year, over 14 million tons of plastic
waste is generated by the United States, of which only 7 percent is recycled.6
This underscores the importance of sustainability and the development of
biobased materials.

This concept of sustainability can be envisioned by a basic understanding
of the dynamics of the global carbon cycle.7 The amount of carbon in the
atmosphere, which exists as carbon dioxide, is kept in balance by arth’s natural
cycles, and it, with the addition of water and sunlight energy, promotes bio-
organic growth through photosynthesis. Over geological time frames this
biomass is fossilized into petroleum, natural gas, and coal—the raw materials
used in the production of polymers, chemicals, and fuels.8 Once consumed,
these fossil resource-derived products release carbon back into the atmosphere
at a much greater rate than that at which it is being consumed. This imbalance
leads to a surplus of atmospheric carbon content, which has been declared to as
the primary cause of global warming.

In a sustainable system, bio-organics are used in the direct production of
polymers, chemicals, and fuels. This can help to mitigate the impact of
petroleum-based materials on the rising carbon content in the atmosphere.
These biobased products, at the end of their life, release the same amount of

carbon dioxide that is required for their growth, therefore helping to equalize the

rate of consumption and emisson.8 The production of petroleum-based plastics
not only results in a disruption of the global carbon cycle caused by a surplus of
greenhouse gases in the atmosphere,9'10 but also contributes to our nation’s
dependence on foreign oil. This realization is becoming a major concern as oil
prices continue to rise and societal demands for a clean and healthy environment
move to center stage. With economical and environmental threats continuing to
surface, greater impetus is focused on new technological advancements.
Materials are being sought to decrease the dependence on imported oil, to
reduce carbon dioxide emissions, and to generate more economic opportunity by
presenting entirely new value-added pathways for agriculture and industry
allowing for major gains in farm income and rural development in the United
States.

The rekindled interest of plastics derived from biological precursors has
been at the forefront of research objectives for the past 30 years. Now more than
ever, with an increased demand,11 efforts are continuously being made by the
United States government to regulate the consumption of fossil fuels and to
promote the use of annually renewable resources. This has stimulated the urgent
development of materials from. biobased feedstocksfz13 According to a market
study by Rapra Technology,14 the global consumption of biobased and synthetic
biodegradable polymers is expected to increase by nearly 20 percent between
the years 2005 and 2010. A major step in this direction was seen in early 2006,
when federal agencies began complying with federal procurement programs set

forth by the 2002 Farm Security and Rural Investment Act to establish biobased

preference programs.15 Many US cities have even begun to place bans on
polystyrene for food packaging and plasticware. Regulations such as these have
compelled plastic manufacturers to incorporate the concepts of sustainability,
industrial ecology, biodegradability, and recyclability into their considerations for
materials design and selection with the aim of producing commercially valuable
and environmentally friendly polymers. The expectation is for the biobased
plastics market to reach 3 billion pounds or more in the next two years.13 Such an
innovation would boost the economics of energy crops and biomass for biobased
polymer and biofuel production, with the long term goal of one day displacing
many of the 280 million gallons of oil currently consumed annually with a
renewable source.16

For the reasons outlined above, there is an emerging market for biobased,
biodegradable polymers that can be disposed of safely and efﬁciently. This
market, driven by rising oil prices and growing health and environmental
concerns, has placed precedence on developing materials from renewable
feedstocks such as corn, sugar beets, sugar cane, wheat, rice, and sweet
potatoes.17 Although, development of numerous novel biobased polymers is
currently underway, economic and engineering barriers may encumber their
success. One of the major issues faced with overcoming these barriers is
meeting the performance requirements of conventionally used plastics. Research
and development of novel materials and improvements upon existing biobased

polymers are needed to enable competition on a comparable price and

performance level and is crucial for the successful implementation of biobased
products into a variety of plastics niche markets.18

Many biobased plastics, including cellulose-based polymers, polylactic
acid, polyhydroxyalkanoates, and starch-based polymers, are increasingly being
investigated.19 In this work, of particular interest is polylactic acid (PLA). PLA has
been receiving considerable attention lately and is viewed by many as part of the
solution to addressing the market need for biobased materials.19'2° This is partly
due to its increasing abundance, but also because it is easily recovered from
plants, relatively inexpensive, and holds great potential to be modiﬁed and
incorporated into polymeric blends and composites with good end uses. The
combination of good physical and mechanical properties along with its
biocompatibility has deemed PLA a key player in the biomedical arena, with
applications ranging from drug delivery systems and arterial grafts to implants,
fracture ﬁxation, and bioresorbable sutures.21 Recent advances, including the
reduction in cost of lactide monomer production and new technological
advancements, have permitted PLA to be introduced into a broader array of
commercial applications.”24 This is especially true in the biodegradable
commodity polymer industry such as in the packaging ﬁeld with single-use,
disposable items and in agricultural mulch ﬁlm production. However, PLA’s full
potential has yet to be exploited. The commercial success of PLA is a function of
its utilization and its utilization relies on its ability to compete with petroleum-

based products by overcoming its inherent drawbacks.

SCOPE OF THE STUDY

PLA is a polymer that has many properties superior to those of other
biobased polymers currently in commerce. On the basis of economic and
environmental considerations, the use of PLA presents a major opportunity for
the development of new biobased products that can be implemented into existing
plastics applications. However, some drawbacks of PLA, namely its low heat
distortion temperature, low melt strength, and brittle behavior, have limited its use
and therefore have incited the desire for improvement. In order for emerging
technologies based on PLA to be forwarded, much needed research remains and
a deeper understanding of PLA is necessary and the ways in which it can be
modiﬁed. Thus, the speciﬁc goal of this work is to develop new PLA-based
materials through compounding, blending, and grafting to help to overcome some
of the current challenges faced by PLA. With the underlying objectives met, the

range of applications to which PLA is suited can be greatly expanded.

STRUCTURE OF THE DISSERTATION

The efforts in this study have been directed toward addressing the current
research need of PLA by various methods of reactive modiﬁcation to develop
new materials based on PLA. In the paragraphs that follow are descriptions of
the contents of each chapter.

Chapter 2 provides the necessary background of PLA as it pertains to the
nature of this work, beginning by introducing some of the basic concepts related
to PLA: its chemical structure, polymerization, and stereochemical differences,

and how they relate to its properties. Processing of PLA and degradation

mechanisms are also described. Details of the experimental methods adopted in
this work are presented in Chapter 3.

The subsequent chapters discuss research pertaining to speciﬁc projects.
The ﬁrst project, which is detailed in Chapter 4, concerns the modiﬁcation of PLA
by the grafting of maleic anhydride (MA) onto its backbone. A study of the impact
of peroxide type on the resultant properties of the maleated product was
undertaken. Speciﬁc attention was focused on minimizing degradation while
maximizing MA grafting yield. Achieving a balance between grafting efﬁciency
and molecular weight enables MAPLA to be used as an interfacial modiﬁer in
PLA-matrix composites, as exempliﬁed in Chapter 5. Chapter 5 speciﬁcally deals
with evaluating the potential of the MAPLA as an interfacial modiﬁer in PLA-talc
composites. The purpose of this work was to investigate the mechanical, thermal,
and microstructural properties and to identify a relationship between microscopic
cause and macroscopic effect in the composites.

Chapter 6 introduces the concept of silane grafting to post-polymerized
PLA. The goal was to use reactive extrusion to achieve radical melt
functionalization of an organo functional silane onto PLA followed by the
moisture-induced crosslinking of the silicon moieties. The aim was to evaluate
the crosslinking efﬁciency and to characterize the resultant crosslinked polymers
on the basis of mechanical considerations, thermal properties, and molecular
characteristics.

The ﬁnal project, in Chapter 7, involves the reactive compatibilization of

PLA with a synthetically-derived biodegradable polymer, PBAT. The objective

was to use a transesteriﬁcation catalyst to promote sufﬁcient compatibilization of
the two polyesters, followed by the production of blown ﬁlm. The degree of
compatibilization was assessed by investigating the thermal, mechanical, and
morphological properties. Finally, Chapter 8 concludes the work with a summary

of the ﬁndings and recommendations for future work.

REFERENCES

1. US Congress, Ofﬁce of Technology Assessment, Biopolymers: Making
Materials Nature's Way, OTA-BP-E-102; Washington DC: US
Government Printing Ofﬁce, 1993.

2. Brydson, J.A., Plastics Materials; Butterworth-Heinemann, 1999.

3. Narayan, R., Biodegradable Plastics, Opportunities For innovation in
BiotechnologY; National institute of Standards & Technology (NIST)
publication, 1993.

4. Stevens, E. S., Green Plastics: An Introduction to the New Science of
Biodegradable Plastics; Princeton University Press, 2002.

5. http:l/www.epa.gov/osw/conserve/materials/plastics.htm#beneﬁts.
6. http:l/www.epa.gov/epaoswer/non-hw/muncpl/pubs/msw06.pdf.

7. Narayan, R., Drivers and rationale for use of biobased materials using
LCA (Life Cycle Assessment), Proceedings, Society for Plastic Engineers
(SPE), GPEC - 2004, February 18-20, 2004, Detroit ,MI.

8. Narayan, R., Plastics from Renewable Resources, GPEC 2005, Global
Plastics Environmental Conference—Creating Sustainability for the
Environment, February 23—25, 2005.

9. Narayan, R., Biobased & Biodegradable Polymer Materials: Rationale,
Drivers, and Technology Exemplars; American Chemical Society
Symposium, 2006.

10. Duncan, M., Journal of Industrial Ecology, 2004, 7, 3,193.

11.US Biodegradable Plastics Market, Freedonia Market Study, January
2008.

12.CLS NRC (Commission on the Life Sciences, National Research Council).
2000. Biobased industrial products: Priorities for research and
commercialization. Washington, DC: National Academy Press.

13.BRDB (Biomass Research and Development Board).2001. Fostering the
bioeconomic revolution in biobased products and bioenergy: An
interagency strategic plan. Washington, DC: BRDB.

14.Smith, R., Rapra Technology, LTD, Woodhead Publishing, 2005.
15. Farm Security and Rural Investment Act, Report 107, 2002.

9

16. Ceres Biotechnology. http/l:www.ceres.net/biofuels.htm|.
17.Sudes, K., Iwata, T., Clean, 2008, 36, 5—6, 433.

18. Mohanty, A.K., Misra, M., Drzal L.T., Natural Fibers, Biopolymers, and
Biocomposites; CRC Press, 2005.

19. Narayan, R., Rationale, Drivers, Standards, and Technology for Biobased
Polymers in Renewable Resources and Renewable Energy: A Global
Challenge; CRC Press, 2007.

20.Goldsberry, C., Plastics: The Bioresin Revolution, Area Development,
2008.

21.Ratner, B.D., Hoffman, A., Schoen, F.J., Lemons, J.E., Biomaterials
Science; Academic Press, 2004.

22.Gruber, P. R. O’Brien, M. Polylactides. NatureWorks PLA, in:
Biopolymers. Polyesters III. Applications and Commercial Products, 1st
ed., Doi, Y., Steinbu"chel, A., Eds.; Wiley-VCH: Verlag GmbH, Weinheim
2002, p. 235—250.

23. Drumright, R.E., Gruber, P.R., Henton, D.E., Adv. Mater., 2000, 12, 1841.

24. Dattaa, R., Tsai, S.-P., Bonsignorea, P., Moona, S.-H. Frank, J. R., FEMS
Microbiol. Rev., 1995, 16, 221.

10

Chapter 2

 

POLYLACTIC ACID BACKGROUND

 

Polylactic acid (PLA) is a biodegradable, aliphatic polyester derived from
the simplest alpha-hydroxy acid, lactic acid (2-hydroxy propionic acid).1'2 The
monomer lactic acid was ﬁrst isolated from sour milk by Carl Scheele in 17803"4
and produced on the commercial level in the late 1800s for food, cosmetic,
pharmaceutical, and agricultural applications.2

Lactic acid can be obtained by the bacterial fermentation of carbohydrates
(corn or potato starch; cane or beet sugar; and cheese whey) or by chemical
synthesis. From an economic point of view, the latter is not practical; therefore,
the former is the preferred route.5'6 Lactic acid exists in two optically active
conﬁgurations, namely as L(+) and D(-) isomers. Both isomeric types are
produced in bacterial systems; however, only the L(+) isomer, which is easily

assimilated during metabolism, is produced in mammalian species.2

11

From a polymer standpoint, lactic acid was of no practical use until it was
later solidiﬁed by self-esteriﬁcation to produce low molecular weight dimers and
oligomers. In 1932, Carothers’ pioneering research led to the production of a low
molecular weight polyester from lactic acid. Then, in 1954 DuPont patented a
process that yielded a high-molecular weight PLA from the cyclic dimer of lactic
acid, lactide. This process entails a two-step method, which will be discussed in
the forthcoming section.7

Due to its chiral nature, lactide exists as three stereoisomers: D-lactide, L-
lactide, and D,L-lactide (mesa-lactide), corresponding to D-lactic acid, L-Iactic
acid and combinations thereof, respectively. Each of the lactic acid isomers and
their lactide monomer structures are shown in Figure 2.1. L-lactide and D-lactide
are chiral enantiomers and have a diastereometric relationship with the achiral
meso lactide.5 PLA can be polymerized from any combination of the three lactide
isomers, producing random copolymers whose stereochemistry is dependent
upon the feedstock type, temperatures, and catalysts used during

production.2'8'9'1o

12

 

    
 

D-Lactic ACId (R) L-Lactic ACICI (S)
H §CH3 H3C \\H
s OH 3 OH
HO HO
O 0
¢ \ o / ¢
0 \\\\\CH3 O
O n
\\\\\CH3 CH3
0 o 0
H30
\\\\\‘ O 0
H30“ 0 H3C
mesa-lactide
O O
D-lactide L-lactide

 

 

 

Figure 2.1: Schematic illustration of the stereochemically different lactic acid

monomers and their corresponding cyclic dimer lactide monomers

PLA Synthesis

The synthesis of PLA for polymer applications is a multistep process. The

production of PLA is performed by either batch or continuous processes and may
proceed by a number of routes. In the simplest route, a basic stepwise
condensation polymerization produces lactic acid consisting of mainly lactyl units.
The method shown in Figure 2.2 can produce one of or both D- and L—lactyl units
in varying ratios. This direct polycondensation method is an equilibrium reaction;
therefore, water must be removed to drive the reaction to the right. In addition,

high temperatures and long reaction times are required. The resulting polymer is

13

a glassy, brittle, low molecular weight PLA with substandard mechanical
properties. In order to obtain a PLA polymer with useful end properties via this
direct route, chain extenders must be added to increase the molecular weight.9
The chain extenders used in PLA are usually bifunctional low-molecular weight
compounds that increase the molecular weight of polymers in a fast and
convenient extrusion reaction. Though the use of chain extenders do contribute

11,12

to additional costs, one major advantage is that most chain extending agents

do not produce byproducts, which could contaminate the resulting polymer.13

0 0 CH3 0
H H H
H
OH I 11 OH
CH3 -H20 CH3 0 CH3

Lactic Acid Low molecular weight PLA

 

 

 

 

Figure 2.2: Synthesis of low molecular weight PLA via a condensation
polymerization reaction
Alternate routes are often taken to obtain a higher molecular weight PLA
and are shown in Figure 2.3. One such method is through an azeotropic
dehydration polymerization step involving ﬁrst, a polycondensation reaction.
Then, an azeotropic mixture of water and solvent are removed from the polymer
mixture.14 Because this method requires the use of organic solvents and entails a

time-consuming melt/solid polycondensation step, it is used infrequently.15

14

The most common, yet more costly method of synthesizing high molecular
weight PLA proceeds via the ring opening polymerization of the lactide dimer.16
Variations of this method have been patented by Cargill,17 Camelot

8 and Ecological Chemical Products.19 Each of which allows for a

Technologies,1
controlled chemistry, therefore controllable and reproducible properties, which in
turn broaden the applications of PLA. In this method a naturally occurring L-Iactic
acid, D-lactic acid, or mixture thereof is polymerized to yield a low molecular
weight PLA prepolymer, which is then catalytically depolymerized through an
intramolecular transesteriﬁcation reaction to produce the dehydrated form of the
cyclic diester lactide. This diester lactide can be present in any of the three
stereofonns. The six-member lactide is ring-opened with a metal catalyst
containing tin or aluminum derivatives. The most preferred catalysts used to yield
high molecular weight PLA are Lewis acid catalysts, such as stannous 2-
ethylhexanoate. The polymerization proceeds by a coordination-insertion
mechanism. During the reaction the Sn(ll) atoms are not covalently bound to the

growing polymer chain at any stage of the polymerization.20 Because of this,

more than one chain may be activated.

15

 

 

 

 

 

 

 

O
0 CH3
H/ OH > HO
CH 3 Condensation O 11 OH

Lactic Acid CH3 0 CH3
8 Low molecular weight PLA
'{g (Mw= 1,000 - 5,000 Daltons)
{D
E) -H20 Depolymerization
C
8
.3 x.
.8. ‘ l
g v 0 010m
0 CH3 0 3H CID O

HOVLiO/KWSYKOH / _ Lactide
RIng—opening
CH3 0 CH3 Polymerization
High molecular weight PLA
(Mw > 100,000 Daltons)

 

Figure 2.3: Synthesis of high molecular weight PLA via azeotropic condensation,
chain extension, and ring-opening polymerization

PLA Properties

Like that of many polymers, the physical properties of PLA are dependent
upon a number of factors, the most important being the stereochemical makeup
of the polymer backbone, which can be controlled by the polymerization of the L-
and D- enatiomeric isomers. The ratio of its stereoisomers directly inﬂuences the
kinetics of crystallization, mechanical properties, and processing conditions.21 Of
particular importance is the effect the optical composition has on the

crystallization behavior of PLA. By adjusting the initial distribution of lactides, PLA

16

can be made either semicrystalline or amorphous. Semi-crystalline PLA polymers
are formed by the syndiotactic arrangement of lactic acid by the polymerization of
pure L,L-Iactide or D,D-lactide. Polymerization of the mesa-monomer produces
an atactic chain because the combination of L- and D-lactide disrupts the
stereoregularity of the polymer, inhibiting crystallization, resulting in an

amorphous material.22

The stereochemical composition of PLA also determines the melting
temperature, rate of crystallization, and extent of crystallization. For example, a
stereochemically pure PLA produced from L-lactide (PLLA) has an equilibrium
melting temperature of approximately 207°C and a glass transition near 60°C.21
However, its maximum practical melting temperature is 180°C with an enthalpy of
melting of 40-50 J/g."'23 Copolymerizing L-lactide with meso-Iactide or D-lactide
can yield polymers that melt at temperatures as low as 130°C. This is because
the addition of meso-lactide or D-Iactide disrupts the stacking order of the L-
lactide chains; it also reduces the obtainable level of crystallinity and the rate of
crystallization, but does not signiﬁcantly affect the glass transition temperature.24
An interesting phenomenon found in equimolar mixtures of L-lactide and D-
lactide is the formation of a stereocomplex. This racemic crystalline mixture
possesses a melting temperature of 230°C, which is approximately 50°C higher

than either homopolymer.25

Crystallizable PLA can be crystallized by the incorporation of nucleating

agents, by strain induction, and by annealing at temperatures between 75°C and

17

the melting temperature.24 It was found that annealing these crystallizable
polymers tend to yield two distinct melting peaks. The origin of multiple melting
has been the focus of many researchers” work. Many theories to explain the
occurrence of multiple peaks can be found in the literature. In the case of PLA, it
has been reported that multiple melting is a result of crystal lamellar
rearrangement during crystallization, with the lower temperature endotherrn
corresponding to the melting of the original crystallites and the higher one to their
subsequent reforming into more perfect, higher melting crystals.24 Additional
studies have determined that the occurrence of double melting endotherrns is
strongly dependent on the heating rate.2°'29 Because PLA crystallizes very
slowly, lower heating rates allow more time for thorough crystallization, usually
resulting in the display of multiple melting endotherms. It is important to note that
the total heat of fusion remains unchanged with variations in heating rate
because the degree of crystallinity does not change during reorganization.30 It
does, however, change with the incorporation of nucleating agents, including talc,
which have been shown to effectively increase the nucleation density thereby

enhancing the rate and degree of crystallization. 2‘

The stereochemical makeup of PLA, along with its crystalline orientation
and molecular weight, also affects the mechanical properties of the polymer.
Table 2.1 shows the mechanical properties of PLLA and PDLA. As shown,
crystallized PLA polymerized solely with L-Iactide has higher tensile and impact
properties than a PLA polymer comprising of D- and L-lactide because of its

ability to crystallize and its higher molecular weight. Annealed L-PLA exhibits the

18

highest mechanical and thermal properties of the three. This is because of
enhanced chain stereoregularity and crosslinking of the crystalline domains,
which lead to respective increases in the tensile and impact strengths.1

Table 2.1: Mechanical Properties of PLA 2

 

 

Annealed
L-PLA L-PLA D,L-PLA
Tensile strength (Mpa) 59 66 44
Elongation at break (%) 7 4 5.4
Modulus of elasticity (MPa) 3750 4150 3900
Yield strength (MPa) 70 70 53
F lexural strength (MPa) 106 1 19 88
Unnotched Izod impact strength (J/m) 1 95 350 1 50
Notched Izod impact strength (J/m) 26 66 18
Rockwell hardness 88 88 76
Heat distortion temperature (°C) 55 61 50
Vicat penetration (°C) 59 165 52

 

Among the other properties of PLA, some of the properties that make it
attractive in packaging applications are its good crease-retention, crimp, and heat
sealability properties, and excellent resistance to oil. It also acts as a good barrier

to ﬂavors and aromas.‘21

PLA Biodegradation

Biodegradable polymers in this context are those “which can be disposed

of in bioactive environments and degrade by the enzymatic action of

”31

microorganisms and one that can be “converted safely into carbon dioxide,

biomass and water.” 32

Polymer degradation mostly occurs through
macromolecular chain scission. In the environment, a variety of chemical,

physical, and biological processes play a role in the degradation of polymers.32

19

Therefore, different degradation mechanisms may be involved, including thermal
activation, hydrolysis, biological activity (enzymes), oxidation, photolysis, or
radiolysis. These mechanisms can occur in combination or in isolation, alongside
other dependent factors such as pH, phase, temperature, exposure, mechanical
stress, and biological activity that affect the polymer degradation process. In
addition, the properties of a polymer are important and can alter the process.33
Some of the properties affecting degradation consist of polymer diffusivity,

porosity, morphology, crosslinking, purity, and chemical reactivity.11

The primary method by which PLA biodegrades is hydrolysis and has
been studied both in-vivo and in vitro.‘°"“37 Typically, degradation in PLA is auto—
catalyzed by carboxylic end-groups, resulting in the random non-enzymatic chain
scission of the ester linkages. This eventually produces monomeric lactic acid
and leads to a reduction in molecular weight with the associated catastrophic
loss in mechanical strength.38 Next, the molecular integrity of the chain
deteriorates. As the molecular weight approaches very low values (Mn~10,000-
40,000 g/mol), microorganisms consume the remaining oligomers, converting
them into carbon dioxide and water. 1'24'38 The degree and rate of hydrolysis are
dependent upon a variety of factors including time, temperature, moisture,
molecular weight, low molecular weight impurities, monomer ratio, crystallinity,
and residual catalyst concentrations,2 all of which can lower the degradation

temperature and promote a faster rate of degradation.

20

Polylactic Acid Processing

The rheological characteristics of PLA permit it to be processed by most
conventional modes of polymer processing, including extrusion, injection
molding, blow molding, and ﬁber spinning.21 Film production, however, has been
shown to be quite difﬁcult due to the low melt strength and instability at elevated
temperatures. Nevertheless, by controlling polymer composition and the addition

of stabilizing agents, PLA can be cast into ﬁlms ranging in thicknesses of 8 pm to

510 pm using standard ﬁlm casting equipment.38

The processing conditions, (residence time, temperature, and mechanical
shear rate) as well as any moisture present in the polymer, govern the properties
of the resultant polymer. Being an aliphatic polyester, the main disadvantage of
PLA is its susceptibility to molecular weight loss due to thermal degradation at
temperatures above its melting point, making the window for processing fairly
narrow. It has been postulated that thermal degradation in PLA results not only
from hydrolysis, but also from intramolecular and intermolecular
transesteriﬁcation, oxidative main chain scission, zipper-like depolymerization,
and pyrolytic elimination.1 Each of these reactions has been reported to be
inﬂuenced by a number of parameters, including moisture, molecular weight,

end-groups, residual catalysts, and hydrolyzed monomers.13

PLA Limitations

PLA has shown to possess some very interesting properties, while some

are already sufﬁcient, others present an opportunity for improvement through

21

modiﬁcation. As mentioned previously, PLA is an aliphatic polyester and
because of its structure it degrades readily, which in some cases may cause
unpredictable performance. High temperature applications of unmodiﬁed PLA are
nearly impossible owing to its low glass transition temperature and low heat
distortion temperature. In addition, its low melting point prevents it from
displacing many commodity plastics. Many packaging applications of PLA are
limited by its brittle behavior and low melt elasticity, both of which prevent it from
being blown into ﬁlm. PLA’s water transmissibility rate is very high, increasing its
ability to undergo molecular fragmentation by hydrolysis. Many of these
drawbacks prevent the more widespread commercial acceptance and have
recently been major topics of ongoing research, some of which will be discussed

in the forthcoming sections.

22

REFERENCES

1. Auras, R., Ph.D. Dissertation, Michigan State University, East Lansing, MI,
2004.

2. Garlotta, D., J. Polm Envir, 9, 2, 2001, 63.

9°

Benninga, H., A History of Lactic Acid Making: A Chapter in the History of
Biotechnology: Springer, 1990.

Bigg, D. M., Adv. Polym. Techn., 2005, 24, 2, 69.
Auras R., Harte,B., Selke, S., Macromol. Biosci., 2004, 4, 835.
Sudes, K., Tadahisa Iwata, T., Clean, 2008, 36, 5—6, 433 — 442.

Scott, G., Degradable Polymers: Principles and Applications; Springer, 2003.

9399'.“

Witzke, 0., Ph.D. Dissertation, Michigan State University, East Lansing, MI,
1997.

9. Nieuwenhuis, J., Clin. Mater.1992, 10, 59.

10.L6fgren, A., Albertsson A.-C., Dubois, P., Jérome, R. J. Macromol. Sci-Rev.
Macromol. Chem. Phys., 1995, 379.

11.Hartmann, M. H., Biopolymers from Renewable Resources, 1st ed., D. L.
Kaplan, Ed.; Springer-Verlag: Berlin Heidelberg: Berlin 1998, 367.

12. Dunsing, R., Kricheldorf, H.R., Polym. Bull. 1985, 14, 491.
13. Klyma, J., Acta Polytech. Scan., Chemical Technology Series, 2001, 282, 47.

14.Enomoto, K., Ajioka, M., Yamaguchi, A., Polyhydroxycarboxylic acid and
preparation process thereof, US Patent 5310865, 1994.

15. Belgacem, M.N., Gandin, A., Monomers, Polymers and Composites from
Renewable Resources, Elsevier, 2008.

16.Gruber, R, Hall, E.S., Kolstad, J.J., lwen, M.L., Benson, R. D., Borchardt,
R.L., US Patent 5142023, 1992.

17.Kricheldorf, H.R., Kreiser-Saunders, l., Boettcher, C., Polymer 1995, 36,
1253.

18.Gruber, P., et al,. Continuous process for manufacture of lactide polymers
with puriﬁcation by distillation, US Patent 5357035, 1994.

23

19. Fridman, I., et al., Lactide polymerization, US Patent 5357034, 1994.

20.0'Brien, W., et al., Integrated process for the manufacture of lactide, US
Patent 5521278, 1996.

21.Drumright, R. E., Gruber, P. R., Henton, D. E., Adv. Mater., 2000, 12, 23,
1841.

22. Dorgan, J., Leherrneier, H., Mang, M., J. Polym. Environ., 2000, 8, 1.

23. Pyda, M., Bopp, R. C., Wunderlich, B., J. Chem. Thenn., 2004, 9, 731.
24.Bigg, D. M., Adv. Polym. Techn., 2005, 24, 2, 69.

25.Tsuji, H., lkada,Y., Macromolecules 1993, 26, 6918.

26.Hiljanen-Vainio, M., Heino, M., and Seppalé, J.V., Polymer, 1998, 39, 865.

27.Jang, W. Y., Shin, B. Y., Lee, T. J., Narayan, R., J. Ind. Eng. Chem., 2007,
13, 457.

28. Rim, P.B., Runt, J.P., Macromolecules, 1983, 16, 762.
29.Younes, H., Cohn, D., Eur. Polym. J., 1988, 24, 765.

30.Nijenhuis, A. J., Colstee, E., Grijpma, D. W., Pennings, A. J., Polymer, 1996,
37, 26, 5849.

31.Gross, R., Kalra, B., Biodegradable Polym. Environ. Sci., 2003, 297, 5582,
803.

32.Narayan, R., Biobased & Biodegradable Polymer Materials: Rationale,
Drivers, and Technology Exemplars; ACS Symposium, Ch. 18, pg 282, 2006.

33.Narayan, R. Doi, Y. and Fukada, Biodegradable Plastics and Polymers;
Elsevier: New York, 1994, page 26.1 '

34. Fukuzaki, H., Yoshida M., Asano, M., Kumakura, M., Eur. Polym. J., 1989, 25,
1019.

35.Sodergérd, A., Selin, J.-F., Nasman, J. H., Polym. Degrad. Stab. 1996, 51,
351-359.

36. MacDonald, R. T., McCarthy, S. P., Gross, R. A., Macromolecules, 1996, 29,
7356.

37. Pitt, C. G., Gratzl, M. M., Kimmel, G. L., Suries, J. Schindler, A.,
Biomaterials, 1981, 2, 215.

24

38.Sodergard, A., Stolt, M., Prog. Polym. Sci., 2002, 27, 1123.

39. Bastioli, C., Handbook of Biodegradable Polymers, Rapra Technology Ltd.;
Rapra Technology, 2005.

25

Chapter 3

 

EXPERIMENTAL TECHNIQUES

 

REACTIVE EXTRUSION

Reactive extrusion is a processing technique that offers the potential to
produce innovative materials in a relatively simple and cost effective manner.
Using an extruder as a continuous melt reactor for the production or the
modiﬁcation of polymers was developed in the mid-1960s,1 however, it has
recently experienced a dynamic growth in materials innovation and has become
the dominant route for polymer blending and compounding as many polymer
companies are moving away from classical reaction vessels.”4 From the
standpoint of polymer modiﬁcation and blending, the use of reactive extrusion
proves to be the most efﬁcient means of polymer processing, exempliﬁed by the

following advantages":5
0 The ability for chemical reactions to take place in the absence of solvents,

therefore, solvent removal or a recycling infrastructure is unnecessary

26

Precisely controlled residence time distributions and temperature proﬁles
The ability to combine several processes such as mixing, reacting, and
shaping all into one step

Processing of high viscosity polymers

Improved surface/volume ratio leading to higher reaction rates

In principle, reactive extrusion consists of creating conditions for speciﬁc

chemical reactions between components during mechanical blending. It is often

performed by the introduction of either a reactive third component with

appropriate functional groups or a catalyst. As a result of the process, the

chemical reaction takes place within the extruder. This process has proven to be

an attractive route to carry out variety of reactions including:

3.6.7.8,9

Bulk polymerization to produce high molecular weight polymers prepared
from monomers or oligomers

Grafting reactions for incorporation of side chains along the polymer chain
Copolymerization through coupling of multiple polymers

Functionalization of polymers to introduce functional end-groups onto the
polymer chain, providing improved chemical or thermal stability,
mechanical, morphological, or optical properties

Controlled molecular weight degradation in which higher molecular weight
polymers are reduced to lower weight through exposure of shear, high
temperature, and a free radical generator

Promotion of a reaction by catalysis

27

In this work, reactive extrusion was employed in the development of PLA-based
polymers through grafting, compounding, and blending. The extruder used was a
Century ZSK-30 twin-screw co-rotating extruder consisting of a screw diameter of
30 mm and a length-to—diameter ratio of 42. The system was equipped with an
extruder driver with a speed control gearbox and an Accurate single-screw
feeder. The speciﬁc conditions of each extrusion run are detailed in its

corresponding chapters.

TITRIMETRIC ANALYSIS

Titration is a chemical analysis technique commonly used to determine the
concentration of an unknown reagent using a standard amount of a known
reagent. By knowledge of simple acid-base chemistry, the quantity of acidic or
basic groups in a substance can be determined. There are a variety of titrations
types. The technique employed in this work uses a visual pH indicator whereby
the endpoint of reaction can be identiﬁed by a color change, denoting the
equivalence point. The volume corresponding to this point can then be related to
the amount of the unknown reagent using a balanced chemical reaction. The
grafting yield of reactive monomers can be quantiﬁed using a calibration curve.
Comparison of the results can provide valuable information on the optimum
parameters for maximum grafting efﬁciency.

In this work, titration was employed to quantify the amount of maleic
anhydride grafted onto the PLA backbone (as will be discussed in Chapter 4). In

the titration, 1 g of puriﬁed, dried MAPLA was dissolved in a 100 mL

28

chloroformzmethanol (3:2 v/v) solution. The carboxylic acid concentration was
determined by a direct alkali titration using 0.1 N potassium hydroxide in
methanol, which had been standardized against a solution of potassium
hydrogen phthalate. The solution was titrated using 1 wt% phenolphthalein as an
indicator. The grafted polymer was completely soluble and did not precipitate
during titration. The percentage of MA grafting was calculated using Equation 3.1
and veriﬁed by utilization of a linear regression calibration curve (y=0.9675x—
0.288; R2 = 0.9996) constructed by using a known amount of reagents. Each
sample was tested in triplicate and average values were reported.

N V
%MA grafting: if” K0” x98.06x100 (3.1)

sample

 

N K 0H and VKOH are the normalIty (mole/equwalent) and volume (In mL) of KOH

solution, respectively. M is the mass (in grams) of the maleated sample.

sample

GEL PERMEATION CHROMATOGRAPHY (GPC)

GPC measures the molecular weight and molecular weight distribution of
a polymer as compared to that of a known standard. In the procedure, a polymer
is dissolved in a solvent and injected through a series of columns consisting of
tightly packed porous polymer beads. As the polymer molecules travel down the
column, the smaller solute particles are able to diffuse into the holes of the
porous polymer. The larger polymer solutes can only enter a fraction of the holes,

if any, and are therefore only slightly retarded as they traverse the length of the

29

column and hence have a faster elution time. The molecular weight moments (Mn

and M) are calculated and used to study of the effect of grafting, crosslinking,
and processing on the molecular characteristics of the polymer. Furthermore, the
distribution as given by the polydispersity index (MW/Mn) can provide information
on polymer chain scission, branching, and crosslinking.

Molecular weight measurements of all PLA-based samples were carried
out in a Waters GPC system equipped with an isocratic 1515 HPLC pump (ﬂow

rate = 1 mL/min), a 2414 refractive index detector, and a 717 autosampler.

Samples were dissolved in tetrahydrofuran in concentrations of 2 mg/mL.
Molecular weights (Mn and MW) and molecular weight distributions (MW/Mn) were

calculated using a universal calibration curve constructed with polystyrene

standards.

THERMAL CHARACTERIZATION

Thermal analysis techniques are used to measure the changes in a
material as a function of temperature. Techniques such as differential scanning
calorimetry, thermal gravimetric analysis, and dynamic mechanical analysis are
used to study transitions representing characteristic chemical reactions. The
results obtained can help to understand material behavior during polymer
modiﬁcation, manufacturing processes, and process optimization, while enabling
improvement in material properties. Given the relative importance of the

employed thermal analysis techniques, an introduction will be provided in the

30

following sections to provide some background and appropriate terminology for

the technical discussions that will follow.

Differential Scanning Calorimetry (DSC)

DSC is a therrnoanalytical tool used to study the thermal transitions based
on chemical or physical characteristics. The technique involves the heating of a
sample and reference pan at the same rate over a speciﬁc temperature range.
The heat differential that is formed by increasing the temperature is measured
and plotted as a function of temperature. When the sample undergoes a physical
transformation, the process is either exothermic or endothermic and is displayed
as a DSC thermogram. From the curve, the heat capacities, melt enthalpies,
transition temperatures, and chemical reactions can be measured and can
provide information regarding crystallization, kinetic processes, and phase

transitions.10 In the context of this work, a TA instrument 2920 differential
scanning calorimeter was used to identify the thermal transitions (T9, Tm, Tc) and
assess the changes taking place. Of particular importance are the effects of
modiﬁcations on the crystallization behavior of PLA. The crystallinity, Zc: of the

materials can be calculated according to Equation 3.2.

AH
Zc— m

— X100 0/0 (3.2)
AH; X ¢PLA

 

where AH", is the enthalpy of fusion (in J/g) and AH; is the enthalpy of fusion

of a PLA crystal of inﬁnite size, whose reported value is 93.7 J/g. ¢pLA is the

weight fraction of PLA in the sample.

31

Thermogra vimetric Analysis (TGA)

TGA provides information on a material’s thermal stability and its fraction
of volatile components by permitting analysis of desorption, degradation, and
decomposition processes under a controlled atmosphere. In TGA experiments, a
sample is heated and the changes in weight that occur are measured. The
measurement is normally carried out in air or in an inert atmosphere and the
weight is recorded as a function of increasing time or temperature.11

The main application of TGA in this work was to determine the
decomposition temperatures of the grafted and crosslinked PLA to assess the
effect of the modiﬁcation on the thermal stability of the polymer. In each of the
projects of this study, a high resolution thermogravimetric analyzer TGA 2950
from TA instruments was used to determine the change in weight of the sample
due to decomposition. The sample was heated from 25°C to 500°C at the heating
rate of 25°C/min. The balance and the samples were purged with nitrogen gas at

rates of 40 ml/min and 60 ml/min, respectively.

Dynamic Mechanial Analysis (DMA)

Dynamic mechanical behavior of materials provides a deeper
understanding of molecular relaxation mechanisms and their relationship with
microstructure. DMA is a technique used to assess the changes in the
viscoelastic response of materials over a spectrum of temperature, time, or
frequency. In polymer blends and composites, dynamic mechanical tests are

very sensitive tools for the determination of thermal transitions.

32

Unlike perfectly elastic materials, polymers including PLA are viscoelastic
and have characteristics of both viscous liquids and elastic solids. In DMA tests,
an oscillating force is applied to a sample at a particular frequency, w and the

displacement, a is measured.12 Upon deformation, some of the energy is stored
(elasticity) and some of it is dissipated as heat (damping).13 The stress and strain
for viscoelastic materials are shown by Equation 3.2 and Equation 3.3,
respectively.
a = 0'0 sin (cot + 6) (3.2)
a = so sin (0H5) (3.3)
Where so is the initial strain, 00 is the initial stress, and 6 is the relative angular

displacement of the stress and strain. From the deformation, it is possible to
measure the stiffness, which can be expressed as an in-phase component and
out-of-phase component corresponding to its storage (E’) and loss moduli (E”),

respectively. Thus, it is possible to deﬁne a complex modulus, E* as:

E * = it) = E' + iE" (3.4)
6‘0)

The two moduli are measured simultaneously and their ratio, tan 6 , is the loss

factor representing the ratio of the energy dissipated to the energy stored. ‘4

t 6=— 3.5
an E, ( >

All of these components can be determined from DMA and infers details of
the polymer structure. In this study, a TA 2980 DMA was used to characterize the
changes in properties that take place as a result of crystallinity, molecular

aggregation, and phase separation in modiﬁed PLA, composites, and blends. For

33

the injection molded products, measurements were run on rectangular bars with
approximate dimensions of 57 mm x 3 mm x 12.7 mm using a dual-cantilever
bending clamp at a frequency of 1 Hz and a heating rate of 10°C/min over a
temperature range of 24°C to 140°C. All measurements were performed and
properties were reported in accordance with ASTM D 4065.

For the blown ﬁlms, specimens were cut along the longitudinal direction
having dimensions of 30 mm x 6.45 mm x 0.1 mm and mounted in a tension
clamp. All ﬁlms (except that of neat PLA) were heated from -50°C to 140°C at a
rate of 2°C/min. Tests were run at an amplitude of 50 um and a frequency of 1
Hz. For both the injection molded bars and the blown ﬁlm specimens, the storage
moduli (E’), loss moduli (E”), and tan 8 (E”/E”) were recorded as functions of

temperature.

MECHANICAL CHARACTERIZATION

The mechanical behavior involves studying the deformation of a material
by applied forces. Determination of these properties helps to derive relationships
between the mechanical behavior and the chemical or molecular structure of
polymers, which can, in turn help to predict and improve the performance of

materials.15

Tensile Testing

The tensile strength of a material is a measure of the ability of a material
to withstand stress under a tensile load. During the test, a tensile specimen is
loaded between two mechanical grips and an axial pull is exerted on the sample

at a ﬁxed speed. The extension is continuously measured as the applied force or

34

load increases. Strain is calculated as the ratio of the change in length to the
initial specimen length (Equation 3.6).

_A1

— 76’ (3.6)

8

where 10 is the initial length and Al is the change in length of the specimen as a

result of testing.
The stress is plotted as a function of strain and the ﬁnal stress-strain
curve, such as the one depicted in Figure 3.1, can help to predict the

performance of materials.

———.__——_

Guts

O'ys ------

Stress (o)

 

p——————-—-—-—.—-‘
———_———--—-—n—_—_

V

I
_————_—————

Guts Eu 8f
Strain (0)

Figure 3.1: Schematic diagram representing a typical tensile test of a material
showing the stress-strain relationship
The relationship between the applied force and the elongation is linear. In
this linear region, the curve obeys Hooke's Law (Equation 3.7), where the ratio of

stress to strain is a constant and:

35

, !

1E“___

0=E£ (3.7)
E is the slope of the line in this region where stress (0') is proportional to strain
(a) and is called the Modulus of Elasticity. The maximum load that can be

sustained by the specimen in tension corresponds to the tensile strength, as
denoted by outs in Figure 3.1.

In this work, the tensile behavior of modiﬁed PLA was determined using a
United Testing System (UTS) Mechanical Tester (Model SFM-20) ﬁtted with a
load cell with a maximum of 1000 lbs (for tensile bars) or 100 lbs (for blown ﬁlm
specimens). Prior to testing, specimens were conditioned at 25°C and 50%
relative humidity for at least 40 hours. The crosshead speed was set at 0.5
in/min. As will be discussed in the forthcoming sections, the tensile behavior of
materials can be drastically affected by the coupling of two or more phases and
thus will be used to provide a correlation between interfacial adhesion and

mechanical behavior.

Impact Testing

Impact tests determine the toughness of a material under high impact
velocity conditions. Speciﬁcally, notched Izod impact tests measure a material’s
resistance to impact by determining the kinetic energy needed to initiate fracture
and continue the fracture until the specimen is broken. Each Izod specimen is
notched to a speciﬁc radius to prevent deformation of the specimen upon
impact. A weighted pendulum is released, swinging through the path where the
sample is ﬁxed. As the sample breaks, energy is absorbed by the sample. The

height the pendulum attains after impact is measured by an indicator on a ﬁxed

36

‘q

scale which reads in joules (ft—lb). The impact strength is then calculated by the
loss of momentum in the pendulum while breaking the sample off at the notch.16
Impact strength can be related to results obtained from tensile tests in that
the area under a stress-strain curve is proportional to the energy required to
break the material. Under the condition that the test is performed at a high
enough rate, the area under the curve is equal to the impact strength of the
material. In many applications, impact strength is a critical measure of service
life. In this work, the Izod impact strength of the PLA-talc composites was
measured using a Testing Machines Inc. (TMI) 43-02-01 monitor/impact
machine. Rectangular specimens cut from Type I tensile bars were notched and

conditioned, according to the ASTM D256 standard.

MELT FLOW INDEX (MFI)

The MFI of a material is a measure of the rate of extrusion of molten
polymers through an oriﬁce of speciﬁc dimensions at prescribed conditions. The
melt ﬂow rate of a polymer is an empirically deﬁned property, which is affected by
the physical properties of the polymer, including molecular stmcture as well as
the measurement conditions.17 Though determining the melt ﬂow rate of a
polymer is not a direct measure of any intrinsic polymer property, it does
correlate, to a certain extent, to the molecular weight and molecular weight
distributions of some polymers.18 As follows, it can be used to differentiate
between grades of a speciﬁc polymer, or in this context, to determine the relative

extent of degradation and/or crosslinking of a polymer as a result of modiﬁcation

37

and processing. A more degraded polymer or a material with fewer crosslinks
has a higher melt ﬂow index as a result of reduced molecular weight and in turn
could result in poorer physical and/or mechanical properties. For the VTMOS-
grafted samples, experiments were carried out using a RayRan Melt Flow
Indexer. All tests were run in accordance with ASTM D1238-90b. The pelletized
samples were melted at 190°C and driven through a capillary die using a 2.16 kg
weight attached to a piston. The mass (in grams) of the extrudate was measured

at 30 second intervals and the MFI was calculated accordingly.

GEL CONTENT AND SWELL RATIO

Gelation

The deterrninatlon of gelation is used to verify the existence of crosslinking
in a polymer. The formation of gel occurs abruptly at the onset of crosslinking. It
is an inﬁnite network of polymer entanglements produced by sufﬁciently high
crosslink concentrations. The gel content of a polymer is a measure of the

insoluble portion of a polymer and can be determined using Equation 3.8:

w
% Gel content = -—f (3.8)

wt

where Wf and Wi are the ﬁnal and initial sample weights, respectively.

Degree of Swelling

The swelling ratio of a crosslinked polymer can be measured by the

amount of solvent uptake in the polymer network. Generally, it is deﬁned “as a

38

”19

 

system consisting of a polymer network swollen with solvent and can be
related directly to the degree of chain crosslinking by Equation 3.9:
Pp
DS = [(W, — Wg)/Wg][ ] (3.9)
Psolvent

where Wg is the average weight of the dry gel component in the crosslinked

sample and W3 is the weight of the gel component swollen at room temperature
for 48 hours in methylene chloride. pp and psolvem are the densities of the

polymer and solvent, respectively.

ELECTRON MICROSCOPY

Scanning Electron Microscopy (SEM)

SEM is a technique used to observe the surface morphology and
composition of a material. From this, polymer compatibility can be determined,
since the appearance of multiple phases would clearly show whether or not
components are miscible. An SEM image is formed when an electron gun
generates a primary electron beam, which under vacuum follows a vertical path
through the microscope. The beam travels through electromagnetic ﬁelds and
lenses to a sample where it generates secondary electrons, backscattered
electrons, and x-rays. Secondary electrons are produced by the interaction
between the primary beam and the sample. They are emitted from a very shallow
region of the material and collected using an Everhart-Thomley detector, which
then conveys the signal to an ampliﬁer to modulate the intensity of the screen

displaying the surface image. Secondary electrons can only come from the top

39

few nanometers of the material, making them topography-sensitive. Thus,
contrast evolves from the surface properties of the sample.20

In contrast, backscattered electrons are primary beam electrons that have
been scattered by the sample nuclei and escape from the surface of the sample.
Their escape depth is greater (at least 1 pm) than that of secondary electrons,
thus, providing compositional contrast. Backscattered electron microscopy
coupled with x-ray microanalysis is an excellent method to determine chemical
composition.21

SEM is often used to observe fracture surfaces, etched surfaces, and
extracted samples of polymer blends and composites to study adhesion
phenomena and the nature of the corresponding phases, which is an important
part of investigating the structure-property relationships of materials. In this
context, SEM is used to determine surface properties and particulate dispersion
in PLA-talc composites, as the images can provide valuable information on the
mode of failure and interesting physical properties of the material, including
mechanical behavior and interfacial characteristics. However, small, dispersed,
and well-adhered phases are often not visible in the SEM because they do not

provide sufﬁcient contrast.

Transmission Electron Microscopy (TEM)

TEM is also a microscopic technique used to study the morphology of a
material and has the ability to detect very small phase sizes (0.2 nm). Similar to
an SEM, a "light source" at the top of the microscope emits the electrons that

travel through a vacuum in the column of the microscope. Using electromagnetic

40

‘I

.5

lenses, the electron beam is magniﬁed and focused into a very thin beam. Unlike
in SEM, the electron beam travels through the ultrathin specimen. Depending on
the density of the material present, some of the electrons are scattered, while
others are absorbed. At the bottom of the microscope the unscattered electrons
reach a ﬂuorescent screen, which give rise to an image of the specimen with its
components displayed in varied darknesses according to their density. However,
in most cases, if the phases do not have atomic number differences, they must

be stained in order to produce adequate contrast.

X-RAY PHOTOELECTRON SPECTROSCOPY (XPS)

XPS also called ESCA (electron spectroscopy for chemical analysis) is a
solid-state spectroscopic technique that allows a surface of a material to be
analyzed quantitatively by measuring the elemental composition, empirical
formula, chemical state, and electronic state of the elements that exist within a
material. The principle of XPS is such that a sample is irradiated with x-ray
photons that are absorbed, causing the emission of photoelectrons. The

photoelectrons have kinetic energies Ekin , given by:22
Ekin =hV—Eb -F (3.10)
where hv is the energy of the photon, Eb is the binding energy of the atomic

orbital from which the electron originates, and F is the work function of the
spectrometer. Each element gives rise to a characteristic set of peaks in the
photoelectron spectrum at kinetic energies determined by the photon energy and

the respective binding energies. Therefore, the presence of certain elements is

41

 

signiﬁed by the presence of peaks at particular energies, with their intensities
corresponding to the concentration of the element within the sampled region.23
XPS can be used to determine the elemental composition of a sample, but only
on the surface. The mean free path for the photoelectrons and the Auger
electrons is relatively short, meaning that the emitted electrons in the solid can
only travel from a certain depth, thus making XPS a very surface sensitive
analytical method with a maximum sampling depth of approximately 10 nm.24

In the VTMOS-grafted samples XPS spectra were obtained using a Perkin
Elmer Phi 5400 ESCA system with a magnesium Ka x-ray source. Samples were
analyzed at pressures between 10'9 and 10‘8 torr with a pass energy of 29.35 eV

and a take—off angle of 45°. The spot size was approximately 250 pmz.

FOURIER TRANSFORM INFRARED (FT IR) SPECTROSCOPY

FTIR is a chemical analytical technique used to determine sample
composition. In FTIR, a sample is excited by infrared radiation, which is absorbed
into the material and converted into energy causing bonds to stretch, contract,
and bend.25 Each chemical functional group absorbs infrared radiation in a
speciﬁc wavenumber range, and as a result, produces a spectrum with bands
corresponding to the wavenumbers that can be correlated to a chemical structure

to identify the functional groups in a sample.

42

LIFJU'W

REFERENCES

599°5‘9’S"

. Utracki, L. A. Can. J. Chem. Eng., 2002, 80, 1008.

Reactive Processing of Polymers, Vol. 2; Canadian Plastics Institute,
Ontario.

Hyun, M.E., Kim, S.C., Polym. Eng. Sci., 1988, 28, 11, 743.

Groeninckx, G., Harrats, C., Thomas, S., Reactive Blending with Immiscible
Functional Polymers: Molecular, Morphological, and Interfacial Aspects In
Reactive Polymer Blending Eds. Baker, W et al.; Hanser: Munich, 2001, 43.

Brown 8.8., Orlando C.M., Encycl. Pol. Sci., vol. 14, p. 169-89, 1988.
Lambla M., Macromol. Symp., 1994, 83, 37.

Carlson, D., MS. Thesis, Michigan State University, East Lansing, MI, 1992.
Gupta, A., MS. Thesis, Michigan State University, East Lansing, MI, 1994.

Xanthos M.; Reactive Extrusion; Hanser: New York, 1992.

10. Billmeyer, F ., Textbook of Polymer Science; Wiley: New York, 1962.

11.Kampf, G., Characterization of Plastics by Physical Methods; Hanser:

Munich, 1986.

12.Menard, DMA: Introduction to the Technique, Its Applications and Theory;

CRC Press, 1999.

13.Young, R., Lovell, P., Introduction to Polymers; 2"° ed.; Chapman & Hall:

New York, 1991.

14.Thermal Characterization of Polymers, 2nd ed.; Turi, E., Ed.; Academic

Press: San Diego, 1991; vol.1.

15. Nielsen, L., Mechanical Properties of Polymers; Reinhold: New York, 1962.

16. Instron Home Page. http://www.instron.us (accessed August 2008).

17.ASTM Standard 1238-01.

18.Bremner, T., Rudin, A., J. Appl. Polym. Sci., 1990, 41, 1617.

19.Sperling, L.H., Introduction to Physical Polymer Science, 3'° ed; Wiley:

Hoboken, NJ, 2001.

43

20.Flegler, S.L., Heckman, J. W., Klomparens, K., Scanning Electron
Microscopy, An Introduction; Oxford University Press: New York ,1993.

21.Sawyer, L., Grubb, D., Polymer Microswpy, 2"° ed; Chapman 8 Hall:
London,1996.

22. Briggs, 0., Surface analysis of polymers by XPS and static SIMS;
Cambridge University Press, 1998.

23.Queen Mary University of London, Department of Chemistry website:
http://www.chem.qmul.ac.uk/surfaces/scclscat5.htm.

24. Ratner, B.D., Castner, D.G., “Electron Spectroscopy for Chemical Analysis”
in Surface Analysis: The Principle Techniques, Vickennan, J.C., John Wiley
& Sons: Chichester, England, 1997.

25.Silverstein, R.M., Webster, F.X., Spectrometric Identiﬁcation of Organic
Compounds, 6‘“ Ed., John Wiley & Sons, Inc.: New York, 1998.

44

Chapter 4

 

EFFECT OF PEROXIDE ON THE MALEATION OF POLYLACTIC ACID

 

INTRODUCTION

Polymer compounding is a well established route leading to the
development of new polymeric-based materials and is attracting more and more
attention because of their expanded applications. Its continued growth is driven
by the ability to easily achieve tailored properties and cost reduction over
traditional unﬁlled polymers. Recently, federal bylaws and increased
environmental concerns have encouraged the development of these materials
from renewable resources. In view of this, biobased polymers, particularly
polylactic acid (PLA) used as composite matrices, have become of recent
commercial interest. These materials have been experiencing rapid growth in a
number of industries, with the expectation of replacing some petroleum-based

plastics. Nevertheless, this expectation cannot be met without overcoming some

45

of the technical hurdles associated with the development of PLA-based
composites.

A major hurdle is the incompatibility between phases. PLA is a
hydrophobic polymer, whereas most ﬁllers are hydrophilic. This chemical
dissimilarity makes interfacial adhesion in a physical blend impossible,
preventing necessary stress transfer from one phase to the other, resulting in
poor performance. Thus, in order to fully maximize the reinforcement effect of the
ﬁller, the requirement of sufﬁcient interfacial bonding must be met. Considerable
efforts to improve the physical and mechanical properties of multiphase PLA-
based materials have gamered some success by the addition of
compatibilizers.“3

Compatibilizers are intermediate components that are preferentially
situated at the interface of two phases. Their use is based on a rather straight-
forward concept. In general, they are multi-functional, bearing distinct segments,
each of which has an afﬁnity to one of the phases. Once added to the system,
their role is essentially to create a bridge between the incompatible phases
through chemical or physical bonding. The result is a reduction in interfacial
tension, enhanced adhesion, and improved dispersion of the ﬁller.

In many cases compatibilizers are produced from polymers that are
grafted or functionalized with reactive molecules, which enable them to interact
with the functional groups of one or both phases. It is common for the

functionalization of a polymer to be performed via a reactive extrusion method,

46

which combines the chemical modiﬁcation step with the processing step into a
single process.“'5

Several functional monomers that have been used in melt grafting
reactions include isocyanates, amines, carboxylic acids, epoxides, oxazolines,
and combinations thereof; however, the most commonly used is maleic
anhydride (MA). MA is a polar monomer that has been extensively studied and
shown to be an effective mode to improve the adhesion between two or more
phases. From a commercial standpoint, it is the most important and oftentimes
the monomer of choice, primarily because of its low cost, its high reactivity, and
the relative ease with which it can be grafted onto polymers in an extruder.° This
ease stems from the dual reactivity that MA possesses, resulting from the free
radical reactivity of the double bond and functional reactivity through the cyclic
anhydnde.

The maleation of PLA involves the grafting of MA onto the PLA backbone
in the presence of an organic peroxide. Having long been performed in the
modiﬁcation polymers, particularly polyoleﬁns7, and more recently in
biodegradable polyesters8'10, the maleation of polymers can lower the interfacial
tension between components to effectively transmit the applied forces from one
constituent to another leading to blends and composites with useful end
properties.”14

The concept of grafting MA onto PLA was ﬁrst introduced in our group by

I 15.16

Carlson et a and was investigated as a matrix material in starch-ﬁlled

composites. Studies conducted on the morphology of the resultant materials

47

' nL'-u—_I'

revealed a substantial improvement in adhesion between MAPLA and starch.
Not long after, other researchers followed suit with respect to the development of
MAPLA-starch blends.17 They expounded upon Carlson’s work by investigating
the mechanical behavior and interphase thicknesses of starch-ﬁlled MAPLA.
Enhanced adhesion was reported, as evidenced by improvements in mechanical
properties and microstructural analyses. Various other studies have since been
performed, most of which conﬁrm property improvements and increased
adhesion in MAPLA-based composites and blends."”22

The literature infers that the mechanism, concentration, resultant
properties, and location of MA grafting are dependent upon several variables
including: grafting technique, reaction conditions, and initiator type and
concentration, making it difﬁcult to pinpoint optimal conditions for this process. In
the present study, MA was free-radically grafted onto a PLA backbone by
reactive extrusion. Two structurally different peroxides were used to initiate the
reaction. The effects of varying the concentrations of the peroxides were
examined. The MA content was also varied. The effects of the variations were
studied by molecular and thermal analyses to determine the optimal

concentrations of monomer and initiator to be utilized for future use as an

interfacial modiﬁer in PLA-based composites.

EXPERIMENTAL

Materials

PLA (Nature Works 30510) was obtained from Cargill Dow (Minneapolis,

MN). Maleic anhydride (MA) (Figure 4.1) was purchased from Sigma-Aldrich

48

Chemical Company. Two peroxides, 2, 5-dimethyI-2, 5-di-(tert-butylperoxy)
hexane (L101) (Figure 4.2) and 3,6,9—triethyl-3,6,9-trimethyl-1,4,7-triperoxonane

(T301) (Figure 4.3), were supplied by Akzo Nobel. Both were used as received.

°U°

Figure 4.1: Structure of maleic anhydride (MA) I

CH3 H3 CH3 CH3 {
I I H2 H2 I I .

Hsc—(l—O—O_(|3—C —C —c|:—o—o— I —CH3 «i
CH3 CH3 CH3 CH3

Figure 4.2: Structure of 2, 5-dimethyl-2, 5-di-(tert-butylperoxy) hexane (L101)

CH CH
H C 2 3
80*0'0
H3CHZC$\ CH3
H30 0'0 CHZCH3
Figure 4.3: Structure of 3,6,9-triethyI-3,6,9-trimethyl-1,4,7-triperoxonane (T301)
Methods

Maleic Anhydride Grafting
MA grafting was performed according to a modiﬁed experimental protocol

by Carlson et al.16 Prior to extruding, the PLA was dried overnight in an oven at
80°C to remove excess moisture. The dried PLA pellets were combined with
varying amounts of MA and peroxide (L101 or T301). The peroxide concentration
employed was varied between 0% and 0.5% while the MA concentration was
held constant at 2%. Likewise, the MA concentration was varied between 1% and

3% at a constant 0.5% peroxide concentration. Each mixture was manually

49

tumble-mixed before being introduced into a Century ZSK-30 twin-screw co-
rotating extruder using a mechanical feeder set at a feed rate of 150 rpm. The

extruder screws have a diameter of 30 mm, with an L/D ratio of 40. Previous

L16 I.23

studies by Carlson et a and Nabar et a concerning MA grafting onto
polyesters conﬁrmed that temperature has an insigniﬁcant effect on the grafting
yield. Therefore, only one processing temperature range was employed for each
peroxide type corresponding to a half-life of 1 minute.

L101 initiated samples: 25/90/145/175/175/175/175/175/170

T301 initiated samples: 25/90/165/190/190/190/190/190/185
For all samples, the extruder screw speed was maintained at 100 rpm and a
continuous ﬂow of nitrogen was purged throughout the extruder. Unreacted MA
and excess moisture were removed by applying vacuum at the vent port. The
extrudate was cooled in a water bath and pelletized downstream.
Fourier Transform Infrared Spectroscopy

FI' IR spectra were recorded using a Perkin Elmer 2000 FTIR

Spectrometer. Thin ﬁlms were obtained by dissolving MAPLA samples in
dichloromethane and solvent casting onto a glass disc. The ﬁlms were dried
overnight under vacuum to remove any unreacted maleic anhydride and to

transform any diacid into the cyclic anhydride form. The samples were scanned

from 4000 to 400 cm‘1

Titration
The grafting yield was determined by titrimetric analysis according to a

slight modiﬁcation of a procedure presented by Nabar.23 Each sample was dried

5O

at 80°C under a vacuum to evaporate any unreacted MA. The grafted sample
was then dissolved in 100 mL of chloroform and 1 mL of hydrochloric acid (1M)
was added to hydrolyze the anhydride groups. The solution was stirred
vigorously for 30 minutes. The grafted polymer was selectively recovered by
precipitation into methanol, followed by ﬁltration, then drying at 80°C. 1 g of
puriﬁed, dried MAPLA was dissolved in a 100 mL chloroformzmethanol (3:2 v/v)
solution. The carboxylic acid concentration was determined by a direct alkali

titration using 0.1N potassium hydroxide in methanol, which had been

if -' -_._ *-e--.-_-1r ‘-

standardized against a solution of potassium hydrogen phthalate. The solution
was titrated using 1% phenolphthalein as an indicator. The grafted polymer was
completely soluble and did not precipitate during titration. The percentage of MA
grafting was calculated using Equation 4.1 and veriﬁed by utilization of a linear
regression calibration curve (y=0.9675x—0.288; R2 = 0.9996) constructed by
using a known amount of reagents. Each sample was tested in triplicate and

average values were reported.

 

N V
% MA grafting: KOH KOH x98.06x100 (4.1)

sample
where N KOH and VKOH are the normality (mole/equivalent) and volume (in mL)

of KOH solution, respectively. M is the mass (in grams) of the maleated

sample

sample.

Gel Permeation Chromatography (GPC)
Molecular weight measurements of PLA and MAPLA were carried out in a

Waters GPC system equipped with an isocratic 1515 HPLC pump (ﬂow rate = 1

51

mL/min), a 2414 refractive index detector and a 717 autosampler. Samples were

dissolved in tetrahydrofuran in concentrations of 2 mg/mL. Molecular weights (Mn

and MW) and molecular weight distributions (MW/Mn) were calculated using a

universal calibration curve constructed with polystyrene standards.

Differential Scanning Calorimetry (DSC)
The melting (Tm), glass transition (T9), and crystallization (Tc)

temperatures, along with the enthalpies of fusion (AH) of the maleated PLA
specimens were investigated using differential scanning calorimetry (DSC). The
samples were heated to 200°C at 20°C/min and held for 5 minutes to erase any
themal history. The samples were then cooled to -30°C at a rate of -5°C/min to
observe crystallization behavior, then reheated to 220°C at 5°C/min.
Crystallization and melting temperatures were obtained from the cooling and
second heating scans, respectively. The degrees of crystallinity of the
composites were calculated according to Equation 3.1.

Then'nal Gravimetric Analysis (TGA)

A high resolution thermogravimetric analyzer TGA 2950 from TA
instruments was used to determine the change in weight of the sample due to
decomposition. The sample was heated from 25°C to 500°C at the heating rate of
25°C/min. The balance and the samples were purged with nitrogen gas at rates

of 40 ml/min and 60 ml/min, respectively.

52

IF" . ...___._.. .-

Statistical Analysis
Statistical analysis was performed using SPSS for Windows. Thermal and

data were compared by two-way analysis of variance (ANOVA) and a Tukey

post-Hoc test. The mean differences were signiﬁcant at the 0.05 level (p s 0.05).
RESULTS AND DISCUSSION

MA Grafting Reaction

The functionalization of PLA by MA is discussed in the current section.
The effect of the reaction conditions as well as the initiator type and
concentration must be taken into consideration to ensure successful grafting.
Selection of a suitable peroxide for grafting MA onto PLA should involve

consideration of the following:
0 Low toxicity
0 Low volatility
0 Appropriate half-life
0 High hydrogen abstraction ability

In addition to satisfying the above criteria, L101 was selected as an
initiator because of its known ability to induce successful grafting of MA on a
variety of polymers as reported in the literature?“25 T301 is a peroxide that is
quite well established in the modiﬁcation of polypropylene, yet relatively new and
unexplored peroxide in terms of being incorporated in polyesters and polymer
grafting. It is a cyclic tri-functional initiator that potentially holds beneﬁts over

conventionally used peroxides. It was chosen in this project because, upon

53

decomposition, it produces a diradical (shown in Figure 4.4) which, like many
other multi—functional peroxides, may have the ability to produce a higher
molecular weight polymer than those of a lower functionality. Another advantage
is its potential ability to impart branching in the polymer leading to improved melt

strength of PLA. 2°

 

H30 CHZCH3

O O

. . A , 'o o 0 0'

O O &

H3CH202\ *CHa ><:O O/ ></
H30 0‘0 CH20H3

 

 

 

Figure 4.4: Proposed decomposition reaction of T301

The proposed reaction is illustrated in Figure 4.5. In the schematic, it is
shown that the mechanism proceeds via two steps. In the ﬁrst step, the organic
peroxide undergoes homolytic rupture producing either a diradical or multiple
monoradicals, depending on the peroxide type. Then, a hydrogen atom from the
PLA chain is abstracted by a radical attack resulting in a PLA chain radical. In
order for the desired grafting reaction to occur, the macroradical reacts with MA,
which is then grafted onto the polymer backbone and forrrrs a branched
macroradical. This macroradical can then either react with additional MA
monomers or undergo transfer with a hydrogen atom on its chain or another PLA
chain to form a new macroradical. This process continues until grafting is

terminated by recombination.

54

 

 

 

RO'
(”3 ('ng (T3010rL101) (“3 CH3
mc—(|3—Qm «WC—g—Om
H
PLA ROH Q
| o
O
v Maleic
anhydﬁde
ii i“: i i”:
mC—C—Qm mC—C—Om
H__ '—
R ° 0
o O 0 0 O

 

 

Aside from

reacts with MA by

Figure 4.5: Proposed reaction for grafting MA onto PLA

grafting, the PLA chain radical faces two possible side

reactions: crosslinking and chain scission.16 In the former, after the PLA radical

an addition reaction and is grafted onto the polymer chain, the

radical then propagates to form a longer MA grafted pendant. This reaction is
undesirable in MA grafting because it consumes the reactive sites, preventing
desired grafting reactions from occurring. In the latter reaction, after the
monomer addition, the skeletal bonds of PLA are broken and the radical is
transferred to a fragmented chain end. Excessive chain scission leads to lower
molecular weight and poor performance in the resultant material. The probability
of each side reaction depends on both the nature of the macroradical and the

nature of the polymer backbone.

55

Although unlikely to occur, an additional side reaction that becomes
relevant in any system involving MA and free radicals together is the
controversial homopolymerization of MA versus grafting of MA. Although this
work will not delve into the speciﬁcs of kinetics of polymerization, it is helpful to
address some of the issues pertinent to MA grafting. Conﬂicting reports by
Russell 27 and Gaylord 28 have spurred much uncertainty of the occurrence of MA
homopolymerization in a variety of systems. Others29 have investigated the
thermodynamics of the possible reaction on the basis of the ceiling temperature
which precluded homopolymerization of maleic anhydride at
temperaturesexceeding 160°C. Furthermore, calculation of reaction ratios
provided evidence that at high reaction temperatures, the reaction favors
hydrogen abstraction over homopolymerization of MA by primary radicals. Thus,
for grafting of MA onto PLA in the present work, analyses were conducted based
on the assumption that homopolymerization would not occur under the

prescribed conditions.

Proof of MA Grafting by Fourier Transform Inﬁared Spectroscopy (F TIR)
The grafting of MA onto PLA using T301 has been attested by FTIR. The

spectra of pure PLA and MAPLA are shown in Figure 4.6. The spectral region
containing the main differences between MAPLA and PLA is highlighted in the
inset plot. Comparison of the spectra within this region reveals a doublet
containing bands at 1759 cm'1 and 1771 cm'1 in the MAPLA sample, where only
one absorption band was observed in that region for PLA at 1755 cm'1

corresponding to its ester group. The doublet in the MAPLA sample is a result of

56

the overiapping of two absorption bands. The band located at 1759 cm'1 is
associated with the carbonyl stretching of the ester group in PLA. The peak at
1771 cm'1 can be assigned to grafted anhydride because ﬁve-member cyclic
anhydrides characteristically exhibit an intense absorption band between 1770
cm'1 and 1792 cm'1 due to symmetric C=O stretching, thus conﬁrming the

grafting of MA onto PLA.

 

 

— MAPLA

 

 

 

 

 

 

 

 

 

1800 1600 1400 1200

cm"

 

 

 

Figure 4.6: FT IR spectra of PLA and MAPLA initiated with T301 verifying MA
grafting

Grafting Efﬁciency and Molecular Weight

In the use of multifunctional polymers for reactive compatibilization, the
degree of functionalization inﬂuences the extent of reaction at the interface,
which, in turn, affects the properties of the resultant multiphase system. The

molecular weight of the compatibilizer is also of prime importance because it is

57

directly proportional to the viscosity of the matrix material, which is a key factor in
the compatibilization reaction and morphological development. The molecular
characteristics and grafting yields of the maleated samples initiated with L101
and T301 were determined. The molecular weights, their distributions, and the
grafting yields were all affected by the MA concentration as well as the initiator
type and concentration. The extent to which they were affected will be discussed

in the forthcoming sections.

Eﬁ'ect of Initiator Concentration on the Grafting Yield of MAPLA

Table 4.1 shows the molecular characteristics as obtained from Gel
Permeation Chromatography (GPC) as well as the grafting yields as functions of

initiator concentration.

Table 4.1: Molecular Weight Data and Grafting Yield as a Function of Initiator

 

 

 

Concentration

Initiator Graft Yield

0
MA /0 W M, (g/mol) Mw (glmol) Mw’ Mn (%)

0* 0 105938 i 4715a 140756 i 70203 1.33 a: 0.13a -
0** 0 96491 :t 1005b 133238 :l: 52508 1.38 :I: 0.07‘11 -
2 0 94805 i 3718b 134119 i 20078 1.4110083 0.0110003
2 0.1 89105 i 1962b 126997 i 549° 1.435004a 0.131002b
2 0.25 72922 :1: 62050 110787 1 3431° 15250.19a 0.191000c

N

0.5 21525 i 3802d 32546 :I: 9331d 1.51:I:0.85a 1.49:I:0.05d

 

0.1 117808115258 1437211143303 1.221014a 0.3710028

2 0.25 9419211626f 116827117236bc 1.241021“ 0.4050006

T301-initiated L101-initiated
N

N

0.5 60080 i 3330b 85930 t 146786 1.431034a 0.70:I:0.00f

 

*Neat PLA; “Extruded PLA

58

The effect of initiator concentration on the grafting of MA onto PLA is
shown in Figure 4.7. The initiator concentration was varied from 0% to 0.5%,
while the MA concentration was held constant at 2%. As was expected, for both
initiators an increase in initiator concentration resulted in an increase in MA graft
yield. The increased degree of grafting can be attributed to an increase in the
number of radicals formed upon the thermal decomposition of the initiator. The
greater the amount of free radicals, the more chain-transfer to polymer, leading
to a higher grafting.

In addition to the concentration of initiator, initiator type also played a role
in the degree of grafting. Comparison of the grafting yields with respect to
peroxide type shows that the samples initiated with L101 displayed much higher
grafting than those initiated with T301. In the L101-initiated samples, a 1.5% MA
grafting yield was seen, while in the T301-initiated samples, a maximum of 0.7%
MA was obtained. It is interesting to note that at low initiator concentrations, the
opposite trend was seen, and the T301-initiated samples had higher grafting than
the L101-initiated samples. This discrepancy can be explained by consideration
of the structures of the peroxides. Upon thermal decomposition, T301 being a
cyclic, multifunctional Initiator generates a radical with two reactive sites, unlike
L101, which produces multiple monoradicals. This means that there are fewer
T301 radicals that are able to partake in the grafting reactions. On the other
hand, the excess diradicals of T301 at higher initiator concentrations may favor

crosslinking or branching as opposed to the desired grafting reactions.

59

 

 

MA Grafted 0%)
O
O)

 

 

 

 

0 0.1 0.2 0.3 0.4 0.5 0.6
Initiator concentration (%)

 

 

 

Figure 4.7: Effect of initiator concentration on MA grafting yield at 2% MA loading

Effect of Initiator Concentration on the Molecular Weight of MAPLA

In the samples prepared with constant MA concentration, increasing the
initiator concentration yielded higher grafting, but it also led to a decline in
molecular weight, as illustrated in Figure 4.8. In other words, the grafting yield
was found to be inversely proportional to the molecular weight. The aliphatic
nature of PLA makes it quite susceptible to hydrolysis during thermal processing,
which is why its molecular weight was determined both before and after
extrusion. However, no signiﬁcant changes were observed between the neat and
the extruded polymer. Thus, ruling out the high temperatures and mechanical

shear as the sole cause of degradation.

As is well known, molecular weight changes in polymers due to the
addition of peroxides are generally associated with chain scission, branching, or

crosslinking reactions. When branching or crosslinking are the dominant

60

reactions, the molecular weight typically increases, but, when chain scission

dominates, the polymer undergoes degradation, lowering its molecular weight.

 

 

  
   

 

 

 

140000
120000 1
*
100000 T301
E 60000 1
E “x
c 40000 . s“
2 ‘s~
20000 - I
o r r I r f
0 0.1 0.2 0.3 0.4 0.5 0.6

 

 

Initiator concentration (%)

 

Figure 4.8: Effect of initiator concentration on the number average molecular
weight of MAPLA at 2% MA loading

The reason for such degradation in the L101 samples is attributed to the
increased concentrations of initiator, which gave rise to a greater number of
propagating macroradicals, allowing chain scission to take place alongside
grafting. This evidently caused the reduction in molecular weight. Conversely, in
the T301 samples, though chain scission may have occurred, branching or light
crosslinking appeared to dominate at low degrees of T301, shown by a relatively
constant molecular weight. At a T301 concentration of 0.5%, a sudden decrease
in molecular weight was seen and attributed to the simultaneous occurrence of
grafting, chain scission, and crosslinking in the system. A small reduction in
molecular weight is not uncommon in MA grafting and is not necessarily a
negative consequence for compatibilizing agents. The surface area of the

interface is, in part, controlled by the molecular weight of reactive polymers. Due

61

to their lower viscosities, lower molecular weight polymers are able to diffuse
faster to the interphase of a multi-component system, where they can
preferentially locate to form the interfacial region to improve adhesion.
Nonetheless, there is a limit below which the molecular weight will have adverse
effects due to the compatibilizer seeping from the interface and mixing with the
bulk polymer.30 These results indicate that an optimum amount of initiator should
be used to achieve maximum grafting of MA with a minimum molecular weight

reduction.

Effect of MA Concentration on the Grafting Yield of MA
Table 4.2 lists the molecular characteristics and grafting yield of MAPLA

obtained as functions of MA concentration. The MA concentration was increased

from 0% to 3% as the initiator concentration remained at 0.5%.

Table 4.2: Molecular Weight Data and Grafting Yield as a Function of MA

 

 

 

 

Concentration

22" L121 Mn (g/mol) Mw (g/mol) Mw/Mn £31023?)

0* 0 10593814715a 140756170203 1.33 1 0.133 -

0** 0 9649111005b 13323815250a 1.38 1 0.073 -
8 0 0.5 9471514297b 1331691112253 1.41 1 0.173 -
a? 1 0.5 74802112694c 1039751220763 1.39 1 0.453 0.151003a
g 2 0.5 2152513802d 3254619331b 1.51 1 0.853 1.491005b
" 3 05 1606214148 2377611833b 1.48 1 0.168 0.741001c
E 0 0.5 1131161492f 140591127636a 1.24 1 0.258 -
3% 1 0.5 8749618976b 126051127669a 1.44 1 0.523 0.441005d
g 2 0.5 60080 1 3330c 85930114678a 1.43 1 0.34a 0.701000e
'- 3 0.5 86776 1 195b 108843 1 55768 1.25 1 0.078 0.411001d

 

' Neat PLA, "Extruded PLA

62

Figure 4.9 presents the MA grafting yield as a function of MA
concentration. The effect of MA concentration on the grafting is much more
complex than the effect of initiator concentration. For both L101 and T301
initiated samples, it was observed that at lower MA amounts the graft content
considerably increased with an increase in monomer concentration and reached

a maximum at 2%; thereafter, it decreased with increasing MA concentration.

 

 

  
   
  

 

 

 

1.6 l
I T‘
1.4 - 1’ 5“ —---L101
A 1.2 - ,’ \\ —o—T301
.\° 1 — ’
V ’1 \\
8 0.8 - I \-
II:
E 0.6 -
o
g 0.4 «
I
0.2 7 I,
'l
O I r T r r
0 0.5 1 1.5 2 2.5 3 3.5
MA added (%)

 

 

 

Figure 4.9: Effect of MA loading on the MA grafting yield at 0.5% initiator
concentration

This effect is similar to results presented by Mani et al.° on the MA grafting
onto polybutylene succinate, who suggested that there is an optimum monomer
concentration or MA—to-peroxide ratio to promote grafting, beyond which the
grafting efﬁciency decreases. This is apparently the case in the current system
as well. At lower MA loadings, the PLA macroradicals reacted directly with MA.
As the MA concentration was increased, termination reactions became prominent

and the grafting content decreased. This suggests evidence of competition

63

between MA and the PLA chains to react with primary radicals generated by

peroxide decomposition.

Effect of MA Concentration on the Molecular Weight of MAPLA

The effect of MA concentration on the number average molecular weight
is shown in Figure 4.10. Without any MA added, both T301- and L101-initiated
samples retained the molecular weight of the neat polymer. The reason for this
retention could be related to the light crosslinking reactions that may have taken
place or to the deactivation of any residual catalyst in PLA by the peroxides,
preventing hydrolysis.31 In this case, the former is expected to have occurred; the
reason for which will be explained in a subsequent section. In the samples
containing MA, the molecular weight decreased as MA concentration was
increased because of hydrolysis. At higher MA concentrations, however, the
molecular weights in the L101-initiated samples began to level off after 2% MA,
whereas the T301-initiated samples experienced a slight increase in molecular
weight at 3% MA. Because grafting is inversely related to the molecular weight,
both the leveling and the ascent in molecular weight in the L101- and T301-
initiated samples—respectiver--can be ascribed to the increase in grafting
reactions. The differences observed between the T301- and L101-initiated
samples can, again, be related to the structural differences of the corresponding

peroxides in this system.

64

 

1 20000

 

 

 

 

 

100000

80000 -
2 60000 ~
33
._- 40000 1
2 —o—T301 ‘\

0 I I I I T r
0 0.5 1 1.5 2 2.5 3 3.5

 

MA added (%)

 

 

Figure 4.10: Effect of MA loading on the number average molecular weight at
0.5% initiator concentration

Thermal Stability of MAPLA

L101-Initiated Samples
Figure 4.11 displays the thermograms of the maleated samples initiated

with varying L101 contents. Each of the samples shows a one-step
decomposition proﬁle. Thermogravimetric analyses of the samples show the
decomposition temperatures to be lower for the maleated polymers than those of
the neat PLA, implying that MA grafting yielded products of lower thermal
stability. Although there are no implications of a causative relationship, the
decreasing trend in thermal stability correlates well with the molecular weight
data. More free radicals in the system promote both chain scission and the
formation of less thermally stable polymer chains. Correspondingly, the presence
of MA in the system also impacted the thermal stability as shown in Figure 4.12.

As expected, the sample prepared with only peroxide was the forerunner in terms

65

of thermal stability due to the fact that the L101 can cause crosslinking, which will
inherently increase the decomposition temperature because of fewer end groups.
Upon addition of MA however, the thermal stability decreased and continued to
decrease as MA concentration was further increased. The decline in
decomposition temperatures can be related to the presence of MA — as the

number of hydroxyl and carboxyl groups increased, so did the ease of thermal

 

 

  
  
 
 

decomposition.
120
- - - - PLA
— — 0% L101
— - - 0.1% L101
—0.25% L101
— — 0.5% L101

  

Weight (%)

 

 

 

300 310 320 330 340 350
Temperature (°C)

 

 

 

Figure 4.11: Effect of L101 on the thermal stability of PLA prepared with 2% MA

66

 

 

 

Weight (%)
o:
O

 

 

 

 

40 —
20 -
0 I I i ﬁr I
290 300 310 320 330 340 350

Temperature (°C)

 

 

 

Figure 4.12: Effect of MA concentration on the thermal stability of PLA prepared
with 0.5% L101

T301-Initiated Samples
Similar to the L101—initiated samples, the thermogravimetric data for the

T301-initiated samples show reduced degradation temperatures as both initiator
and MA concentration are increased. In Figure 4.13 the only sample that
displayed improved thermal stability is the one without MA, conﬁrming that T301

also may lead to crosslinking of the polymer chains.

67

 

 

 

Weight (%)

 

 

 

 

 

310 320 330 340 350
Temperature (°C)

 

 

 

Figure 4.13: Effect of MA concentration on the thermal stability of PLA prepared
with 0.5% T301

At a constant MA concentration (Figure 4.14), the sample initiated with
0.1% T301 had the highest thermal stability, lending to the assumption that while
working alongside crosslinking reactions, chain scission became more
predominant as the initiator concentration increased. Overall, the
thermogravimetric data, in conjunction with the previously discussed molecular
weight data, clearly conﬁrm the fact that molecular weight is correlated with

thermal stability, showing that the lower molecular weight samples are less

thermally stable.

68

 

 

      

 

100 - - - -PLA
_ ___,__ ............... .._‘ _. — 0%T301

80 . ‘ l... - -0.1%T301

0:; (‘1‘ l —— 0.25% T301

*3 60 _ ﬂ '— — 0.5% T301
a) l
'5 '1‘ n
3 40 ~ \‘ I
x‘ '
20 ~ I‘

0 . ' ----- ‘

 

 

 

320 325 330 335 340 345 350 355
Temperature (°C)

 

 

 

Figure 4.14: Effect of T301 concentration on the thermal stability of PLA prepared
with 2% MA

Differential Scanning Calorimetry

DSC was performed on the MAPLA samples to investigate the effect of
maleation on the thermal behavior of PLA. Thermal properties of the maleated
samples as functions of L101 are shown in Tables 4.3. A reduction in the glass
transition temperature was seen; however, no statistically signiﬁcant changes
(p 2 0.05) were observed in the melting temperature of the sample extruded with
MA in the absence of L101. The crystallinity also increased two-fold — from 16%
in the extruded polymer to nearly 35%. This is due to the increasing acid groups
induced by the MA addition acting as plasticizers for the bulk polymer. This

increased chain mobility and permitted the ease of chain stacking.

By comparing the DSC results of the maleated and unreacted samples, it

was observed that a small amount of peroxide had no effect on the glass

69

transition, crystallization, or melting temperatures (p 2 0.05). However, in the
samples initiated with a greater amount of L101, the temperatures slightly
decreased while the crystallinity increased, which was due to the increased

molecular motion of the lower molecular weight grafted samples.

Table 4.3: DSC Data of L101-initiated MAPLA with Varying L101 Concentrations

 

 

(yo 0/0 0 o o o 0

MA L101 Tg( C) Tc( C) Tm,1 ( C) Tm,2( C) /°Xm

0 0 58.1110" 106.91318 146.51158 153.711 .23 16111.1"1
2 0 51510.7" 101.512.63 140.0109" 150.1106" 34.7146"
2 0.1 56912.6a 112111.83 147.7106a 154.5103a 22.811 .2c
2 0.25 55,911.23 107.3114a 146.5100a 154.511 .03 28710.3"
2 0.5 51.9103" 110.5143" 140.8125" 144.6126" 30811.5"

 

Different letters within the same column represent signiﬁcant differences at the 95%
conﬁdence level.

The DSC data for the maleated sample in the absence of MA are shown in
Table 4.4. It is shown that the addition of L101 increased the crystallinity of PLA.
As discussed previously, the molecular weight of this sample remained fairly high
and the thermal stability was improved. Therefore, under the assumption that
branching had occurred concurrently with grafting it was expected that the
polymer would have a lesser degree of crystallinity.15 However, this was not the
case. The increased crystalline fraction could be a result of microgel particles
acting as nucleating agents in the bulk polymer promoting crystallization. In the
sample prepared using 1% MA, the Tc was 11% higher and experienced a 15%
increase in crystallinity compared to the pure extruded polymer. In the 2% MA
sample, a decrease in the melting and glass transition temperatures was

observed with a slight increase in crystallinity. From the information presented in

70

the previous section, it should be recalled that this sample had the highest
grafting percentage. The 2% MA sample also corresponds to that which had the
lowest number average molecular weight, which accounts for the increased chain
mobility and consequently lower glass transition and melting temperatures and a
higher crystallinity. At 3% MA, the glass transition and melting temperatures were
equal to those of the 1% MA sample, but the higher MA concentration induced a

higher crystalline fraction due to the severe degradation of the polymer chains.

Table 4.4: DSC Data of L101-initiated MAPLA with Varying MA Concentrations

 

"/o %

 

MA L101 Tg(°C) Tc(°C) Tm,1(°C) Tm,2(°C) ‘Voxm

0 0 58.1110" 106.9131" 146.5115" 153.7112" 16.1111"
0 0.5 58.1124" 102.5118" 144.3108" 152.5114" 22.7108"
1 0.5 57812.1" 114.2116" 146.4104" 150.8106" 26110.7"
2 0.5 51910.3b 110.5148" 140.8126" 144.6126" 30.8115d
3 0.5 54811.2" 108.111.7"" 146.8123" 154.4109" 40.4112"

 

Different letters within the same column represent significant differences at the 95%
conﬁdence level.

The effect of T301 concentration on the thermal behavior is shown in
Table 4.5. In the absence of MA, the 0.5% T301 sample, like the 0.5% L101
sample also experienced a sudden increase in crystallinity compared to the
unmodiﬁed PLA and is also attributed to the formation of microgels that nucleate
PLA, leading to an increase in crystallinity. Of the maleated samples, that which
was prepared with 2% MA and 0.5% T301 showed signiﬁcant differences from
the others, with twice the crystalline content than the other samples with lesser
amounts of T301. This increase corresponds to increased polymer degradation

resulting from MA grafting.

71

Table 4.5: DSC Data of T301-initiated MAPLA with Varying T301 Concentrations

 

%

 

%MA 1301 T900) 01°C) Tm,1(°C) Tm,2<°C) %xm
0 0 58.1110" 106.9131" 146.5115" 153.7112" 16.1111"
2 0 51510.7" 101.5126" 140.0109" 150.1106" 34.7146"
2 0.1 58.2109" 113.8113" 144.2104" 154.2106" 13.9112"
2 0.25 57012.1" 119.8144" 149.1123" 154.7118" 12013.3"
2 0.5 55.7104" 115.6126" 144.9119" 154.1109" 25.7120"

 

Different letters within the same column represent significant differences at the 95%
conﬁdenceleveL

DSC results of the samples maleated with a constant concentration of

T301 are shown in Table 4.6. As is evident, no signiﬁcant changes were

observed in the temperatures or the crystallinities as a result of maleation, fueling

the assumption that the initial concentration of radicals has more of an inﬂuence

than MA in terms of what dictates the molec‘ular behavior of the polymer chains

in this system.

Table 4.6: DSC Data of T301-initiated MAPLA with Varying MA Concentrations

 

 

J; 7% Tg(°0) Tc(°C) Tm1(°C) Tm2(°C) %xm
301

0 0 58.1110" 106.9131" 146.5115" 153.7112" 16.1111"
0 05 57.9106" 107.3117" 146.1104" 153.6106" 35413.7"
1 0.5 58.5115" 119.2132" 149.0126" 155.0119" 22.4107"
2 0.5 55.7104" 115.6126" 144.9119" 154.1109" 25.7120"
3 0.5 57.1107" 120.3108b 149.2113" 154.9120" 28.1103"

 

Different letters within the same column represent significant differences at the 95%
conﬁdenceleveL

CONCLUSIONS

MA was successfully grafted onto PLA through reactive extrusion and the

products were characterized on the basis of molecular weight, graft yield,

mechanical properties, and thermal stability. Two different peroxides T301 and

72

L101 were used to initiate grafting and the chemistry of peroxide initiation was
elucidated. Results on the basis of grafting, molecular weight, and thermal
behavior were compared. The results revealed that the samples that were
initiated with L101 resulted in a higher overall grafting efﬁciency, while those
initiated with T301 underwent the least degradation. The differences were
attributed to the structural differences of the peroxides. Correspondingly, an
inverse relationship between the degree of grafting and molecular weight was
established. It was also found that regardless of peroxide type, there is an
optimum ratio of peroxide to MA (0.5:2) that promotes maximum grafting, beyond
which degradation reactions become predominant. The DSC results indicate that
the crystallinities of the samples were in accord with the molecular weight
measurements, with the exception of the samples containing only peroxide. In
those samples the increase in crystallinity was attributed to microgel particles
acting as nucleating agents. Therrnogravimetric data revealed that the thermal
stability of the samples, for the most part, decreased with the degree of
maleation and the samples initiated with T301 being more thermally stable
overall. Moreover, it was concluded that there must be a balance between
grafting and degradation reactions in MAPLA polymers for their use as effective
interfacial modiﬁers. Furthermore, an optimum MAPLA composition with a high
enough molecular weight with sufﬁcient grafting can be used to promote
adhesion between different phases in composites and blends, as will be seen in

Chapter 5.

73

REFERENCES

1. Thermotolerant Starch-Polyester Composites and Methods of Making the
Same, Nie, L., et al., Patent WO/2007/050560.

Park, J.W., Im, S. S., Polymer Eng.Sci., 2000, 40, 2539.
Ke, T., Sun, X., J. Appl. Polym. Sci., 2001, 81, 3069.
Carlson, 0., MS. Thesis, Michigan State University, 1995

WPPN

Maliger, R.B., McGlashan, S.A., Halley, P.J., Matthew, LG, 2006. Polym.
Eng. Sci., 46, 248.

S”

Vasile, C., Handbook of Polyoleﬁns; CRC Press: New York, 2000.
7. Kamani, R., Krishnan, M., Narayan, R. Polym. Eng. Sci., 1997, 37, 2,477.

8. Mani, R., Bhattacharya, M., Tang, J., J. Poly. Sci., Part A: Poly Chem, 1999,
37, 1693.

9. John, J., Tang, J., Yang, 2., Bhattacharya, M., J. Poly. Sci. Part A: Poly.
Chem., 1997, 35, 1139.

10.Arbelaiz, A., Fernandez, B., Valea, A., Mondragon, I., Carbohy. Polym.,
2006.64.224.

11.0. A. Correa, C. A., Razzino, C. A., Hage, E., J. Thermoplas. Comp. Mater,
2007, 20, 323

12.Qiu, W., Zhang, F., Endo, T., Hirotsu, T., Polym. Comp., 2005, 26,448.

13. Lai, S.-M. , Yeh, F.-C., Wang, Y., Chan, H.-C. , Shen, H.-F., J Appl. Polym.
Sci., 2003, 87, 487.

14. Kamani, R., Krishnan, M., Narayan, R., Polym. Eng. Sci., 1997, 37, 476

15.Dubois, P, Narayan, R., Macromol. Symp. 2003, 198, 233.

16. Carlson D., Nie, L., Narayan, R., Dubois, P., J. Applied Poly. Sci., 1999, 72

17.Sun X., Zhang, J.-F., Biomacromol, 2004, 5, 1446.

18.Plackett, D., J. Polym. in the Environ., 2004, 12, 131.

19. Febrianto, F., Yoshioka, M., Nagai, Y., Tahir, P., Syaﬁi, W., Shiraishi, N., J.
Bio. Sci, 2006, 6, 555.

74

20.Peterson, L., Oksman, A. Mathew, P., J. App. Polym. Sci, 2006, 102, 1852.
21.Huneault, M., Li, H., Polymer, 2007, 48, 1, 270.

22.Avella, M., Bogoeva-Gaceva, G., Buiarovska, A., Errico, M., Gentile, G.,
Grozdanov, A., J. App. Polym. Sci., 2008, 108, 6, 3542.

23. Nabar, Y., Raquez, J. M., Dubois, P., et al. Biomacromolecules, 2005, 6, 2,
807. '

24.Elf Atochem S.A., Graft polymers with controlled viscosity US Patent
5945492, 1999.

25.Bettini, S., Agnelli J., J. Appl. Polym. Sci., 1999, 74, 256.

26.Fityani-Trimm, S., Dhib, R., Penlidis A., Macromol. Chem. Phys., 2003, 204,
436.

27. Russell, K.E., J. Polym. Sci. Part A: Polym. Chem., 1995, 33, 555.

28.Gay|ord, N., Mehta, R., Mohan, D. R., Kumar, V., J. Appl. Poly. Sci. 1992,
44, 1941.

29.Kellou M, S., Jenner, G., Eur. Polym. J., 1976, 12, 883.

30.Jerome, R., Pagnoulle, 0., Key Role of Structural Features of
Compatibilizing Polymer Additives in Reactive Blending In Reaction Polymer
Blending, Eds. Baker, W., Scott, 0., H0, G.-H.; Hanser: Munich, 2001, 83.

31.Sodergard, A., Selin, J.-F., Niemi, M., Johansson, C.-J., Meinander, K., US.
Patent 6559244, 2003.

75

Chapter 5

 

MALEATED POLYLACTIC ACID (MAPLA) AS AN INTERFACIAL MODIFIER

IN PLA-TALC COMPOSITES

 

INTRODUCTION

PLA-based composites are becoming increasingly important, providing the
advantage that one can, In principle, tailor properties to be suitable for speciﬁc
applications. Whether organic or inorganic in nature, ﬁllers can be either ﬁbrous
or particulate, all of which have been incorporated into PLA with the aim of
driving down costs and increasing performance capabilities.1 Natural ﬁllers have
been receiving the most interest.2 An example of this is work by Oksman et al.",
in which the mechanical properties of a ﬂax ﬁber-reinforced triacetin—plasticized
PLA were studied and the results were compared to those of polypropylene (PP)-
ﬂax composites. In that work, promising mechanical properties were reported - a

50%higher tensile strength of PLA with the addition of ﬁbers, as compared with

76

that of the PP-based composites. In an attempt to overcome the brittle behavior
of the composite a plasticizer was added. Resultantly, an Increase in elongation
at break was found, though at the expense of the tensile strength and modulus.
Furthermore, in their study of the morphology, it was concluded that an
optimization of the adhesion at the interface must be met in order to further
improve the mechanical properties. In another study on the thermal and
mechanical behavior, Serizawa et al.4 produced kenaf reinforced PLA-matrix
composites and found that the addition of kenaf ﬁbers increased the
crystallization, thus resulting in signiﬁcant increases in heat distortion
temperature and modulus. However, no improvement in impact strength was
seen with kenaf ﬁbers used as-received. Nevertheless, by eliminating the
crushed ﬁbers and using only the longer ﬁbers in conjunction with a single-screw
extruder (as opposed to a twin-screw extruder), the impact strength was
improved, supporting the well-known theory that composite behavior is very
much inﬂuenced by the type of ﬁber used and by the processing conditions
employed. In a study by Huda 5, the incorporation of wood ﬁbers into PLA
improved the mechanical and thermomechanical properties. Generally, it has
been the consensus that the use of natural ﬁbers can provide a tremendous
increase in both the tensile strength and modulus, however, only few reports
show signiﬁcant improvement in the elongation at break or impact properties
without the use of a plasticizer.

In order to increase the thermal resistance of PLA, researchers have

incorporated into it, starch," clay,7 carbon ﬁbers," and glass ﬁbers,9 each with the

77

aim of increasing crystallization. K9 and Sun" studied the crystallization kinetics
of the PLA/starch composites at varying isothermal crystallization temperatures
and found that the starch addition was effective in increasing the crystallization
rate. Mohamed et al.10 blended PLA with wheat gluten and studied the thermal
characteristics of the composites. They found that the gluten addition led to
decreased crystallization, which they attributed to a high degree of interaction
between PLA and gluten.

Urayama and coworkers11 studied the effects of particle-type and whisker-
type ﬁllers on the thermal and mechanical properties of PLA. The ﬁllers of
interest were calcium silicate, potassium titanate, and aluminum borate, none of
which had any signiﬁcant effect on the glass transition and melting temperatures.
However, the composites did show decreased crystallization temperatures as
compared with the neat polymer. Mechanical data showed that, in general, the
whisker-type ﬁller is much more effective than the particle-type with respect to
matrix reinforcement, with the highest tensile strength being associated with the

potassium titanate and aluminum borate whiskers.

A number of inert ﬁllers, such as talc and calcium carbonate, have also
made their way into PLA-matrix composites for similar reasons as those outlined
above, as well as to reduce cost and improve mechanical properties. Thakur et
al.12 found that talc can act as a nucleating agent for PLA and increases the
number of spherulites during crystallization. Kolstad13 conﬁrmed this ﬁnding and
concluded that the addition of 6% talc gave a 500-fold increase in the nucleation

density. Huda et al.14 incorporated as much as 30% talc into PLA and reported a

78

decrease in tensile strength, but a 92 % increase in tensile modulus. In that work,
the heat distortion temperature was reported to have increased by 24 degrees.

As found in the literature, a number of researchers have produced PLA-
based composites with desirable properties, however, the most important issue
still must be addressed—the compatibility between the matrix and the ﬁller. Based
on a review of the published literature, it is clearly indicated that a simple
uncompatibilized mixture of PLA and ﬁller, regardless of ﬁller type, is insufﬁcient
to produce composites with good end properties. In fact in many cases, simple
admixing can deteriorate the existing properties of the matrix material as any
addition can dismpt the homogeneity and act as a stress concentrator.
Furthermore, the nonpolar nature of PLA makes it difﬁcult to react with
reinforcements containing polar moieties such as ﬁbers, talc, and starch and
assuch, composites with these reinforcements display poor dispersion, a lack of
adhesion, and reduced properties. This presents a major disadvantage to PLA
matrix composites provided that they are unable to reach their full performance
potential. Consequently, the use of coupling agents or compatibilizers is essential
to introduce chemical bonding across the ﬁller/matrix interface.

Substantial experimental effort has been made in an attempt to improve
the phase interaction in PLA-based composites by reactive compatibilization.
This is typically done by modifying either one or both of the composite
constituents by grafting or by the incorporation of a third component — a coupling
agent that acts to reduce the interfacial tension between the phases and

improves the dispersion of the ﬁller in PLA.

79

A well-known compatibilizer, maleic anhydride-grafted PLA (MAPLA), as
mentioned in Chapter 4, can be used as an interfacial agent in many composite
systems. When added into a mixture, it can preferentially reside at the interface
to act as a bridge between binary phases to improve the interfacial adhesion,
which results in a ﬁner dispersion and more stable morphology.15

The use of MAPLA has been exploited in PLA composites reinforced with
inorganic and natural ﬁllers alike. Plackett‘", for example, studied the use of
MAPLA in PLA-based composites separately ﬁlled with wood ﬁber, nanoclay, and
jute ﬁber mat. He showed enhanced adhesion in the PLA/wood ﬁber system, but
no improvement in the latter two. Other PLA-based composites ﬁlled with
cellulose powder, corn starch, and wood ﬂour have been compatibilized with
MAPLA and were reported to have shown improvements in the tensile behavior
and morphology.17 Petersson et al.18 reported improved phase interaction and
mechanical properties from using MAPLA as a compatibilizer in layered-silicate
nanocomposites.

Based on the published literature, MAPLA has been shown to be an
effective method to compatibilize PLA-based composites compounded with
various ﬁllers, however, no studies have been performed on the coupling effect in
a talc ﬁlled system. In this work, talc was chosen as the ﬁller not only because of
its low cost, but also because of its known nucleating effects in PLA. Its structure
is a trioctahedral phyllosilicate with the simpliﬁed molecular formula,
M938i401o(OH)2. Although talc comprises mainly of a hydrophobic face, its edge

surface is composed of approximately 10% hydroxyl groups, which are available

80

for reacting with hydrophilic moieties, such as maleic anhydride to form covalent
bonds.19 Thus, this project aims at investigating the effect of maleic anhydride
functionalized PLA (containing 0.7% MA) as an interfacial modiﬁer in PLA-talc
composites. The concentration of MAPLA was varied from 3 to 10% and the
composites were studied on the basis of mechanical, thermal, and morphological
properties. In the sections that follow, basic theory of composites and interfacial

phenomena will be discussed as it pertains to the context of this work.

THEORETICAL BACKGROUND
Broadly deﬁned, polymer-matrix composites are materials having two or
more constituents of different phases: (1) the polymer matrix and (2) the
reinforcing species.20 The main functions of the polymer matrix are to hold the
reinforcements together in the desired orientation and location, and to transfer
the distributive stresses.” The second phase, the reinforcement, can be
classiﬁed into two groups: ﬁbrous and particulate. Fibrous ﬁllers are
characterized by their large aspect ratios, whereas particulate ﬁllers can be
spherical, cubic, tetragonal, or any other shape. The geometry of the ﬁller has
great bearing on the strengthening mechanism.22 While both types of ﬁllers share
the stress load with the matrix, the particle type does so to a lesser extent and is
not known to generally strengthen the material. Particulate ﬁllers do, however,
stiffen the matrix and improve resistance to creep and oxidation. In general,
properties of composites are determined by:23
c The properties of the ﬁller

o The properties of the polymer matrix

81

‘1) .n_ D-ua

o The volume fraction and volume fraction distribution
0 The interaction between the matrix and the ﬁller

o The orientation of the ﬁllers in the composite

Rule of Mixtures

Knowledge of the relative proportions of a composite’s constituents along
with their respective material properties is sufﬁcient to predict the mechanical
behavior of the composite. A widely used model for predicting the ultimate
strength and modulus of the composite in tensile loading is the rule of mixtures.
The upper-bound approximation (Equations 5.1 and 5.2) is based on the

assumption that the matrix and reinforcement are parallel elementszz‘l'25

O'C =0'mll—Vf)+k0-0'fo (5.1)
Ec =Em(l—Vf)+kEEfo (5.2)

Under the assumption that the matrix and the ﬁller are arranged in series, the
lower bound approximation of the strength and modulus can be estimated by
Equation 5.3 and Equation 5.4: 2"

1 .
— (I—Vf)/0'm+kO-Vf/0'f

 

(5.3)

0c

1

Ec = (5.4)
(l—Vf)/Em+kE Vf /Ef

 

where 0c ,am, and 0' f are the tensile strengths of the composite, matrix, and
ﬁller, respectively. EC ,Em and E f are the respective Young’s moduli of the

composite, matrix, and ﬁller. Vf is the volume fraction of the ﬁller and kg and

82

k E are constants corresponding to the type of ﬁller and its interaction with the

polymer matrix.

Adhesion Theory

It is necessary to understand that the mechanical performance of a
composite is not only a function of the individual components, but also relies on
the ability to effectively transfer stress through the bond and its distribution
throughout the interface.26 The interface is of utmost importance in realizing the
degree to which a composite’s potential strength is met and maintained. In order
for a composite system to obtain its full potential, strong adhesion at the interface
must exist. While the speciﬁcs of adhesion are not the focus of this work, an
elucidation of the various coupling mechanisms will be helpful in understanding
the complex nature of adhesion in the development of composite materials. The
adhesion between the matrix and reinforcement is what dictates the ultimate
performance, as the level of adhesion is directly correlated to resultant
properties. Lacking sufﬁcient adhesion at the interface prevents the necessary
stress transfer between the components. Having an adhesion that is too strong,
on the other hand, limits the energy absorption mechanisms by causing all of the
energy to be focused on a small region of polymer, promoting the ease of crack
propagation and premature brittle failure under stress.27 Therefore, an optimum
adhesion level is necessary at the interface for sufﬁcient mechanical properties.
There are ﬁve primary mechanisms in studying the principles of adhesion”:

. Adsorption and wetting

o Interdiffusion

83

. Electrostatic attraction
. Chemical bonding
0 Mechanical interlocking
These ﬁve mechanisms (whether in isolation or in combination) play a
signiﬁcant role in the adhesion in composite materials as well as those which
have relevance to composite materials will be further discussed in the following
subsection.
Adsorption and Wetting
Thermodynamics provide the basis for the adsorption theory in the study
of ﬁller-matrix surface energetics. Under the assumption of thermodynamic
equilibrium, the following equation was proposed by Young to describe the
relationship between the solid surface free energy and the liquid surface free

energy (Equation 5.5): "9

7SV = 7L5 + 7LV 6089 (55)
where 6 is the angle between the liquid and solid surfaces and l’SV» my, and
7L5 are the free energies of the solid-vapor, liquid-vapor, and the liquid-solid

interfaces, respectively. The precondition for achieving adhesion is wetting. In
order to effectively wet a substrate, the surface energy of the substrate must
exceed that of the liquid resin. Additionally, the resin must initially be of low

viscosity and form a contact angle less than 90° (Figure 5.1)."0

84

  

Liquid
t
7SV 7L5 Substrate

    

 

Figure 5.1: Schematic diagram of a contact angle and its surface tension
components.

Liquids that form contact angles greater than 90" are called non-wetting
liquids. However, if a droplet is not formed, 19. the contact angle is 0",
spontaneous wetting occurs and Equation 5.5 will not hold true. Rather, the
equilibrium reaction can be written as Equation 5.6:

My S l’SV - 7L3 (5-6)

This inequality further implies that in order for spontaneous wetting and
spreading to occur the liquid free energy must be lower than that of the
substrate.”

After wetting is achieved the next essential step to acheive an effective
interphase is the adhesion, which is also governed by the surface energy and
interfacial tension of each of the multiple phases. Adhesion at the interface
allows stress to be transferred across the surfaces of the multiple phases, which
is important to improve physical and mechanical properties of the composite

system. The work of adhesion, Wa , is the maximum force exerted when the two

adhering materials are separated. If only secondary forces are acting, then the

work of adhesion can be expressed by the classical Dupre equation: 3"

Wa = 7LV + 75V - 75L (5-7)

85

Combining Equations 5.5 and 5.7 gives the Young-Dupre equation
(Equation 5.8):

Wa = 7LV(1+ cos 9) (5.8)

When 6 = O, the work of adhesion is at a maximum and the following

relationship exists between the surface energies:
75V "l’SL =7LV (5-9)

This equation implies that in order to have sufﬁcient adhesion, much like
for adequate wetting, the surface energy of the solid must exceed that of the
liquid. Although these thermodynamic relationships are derived for use in
equilibrium conditions, the knowledge of these mechanisms can help to
understand wetting and spreading in compatibilizing polymer composites.
Chemical Bonding

Another type of bonding that is relevant to this work and must be
considered is chemical bonding, which invokes the formation of covalent, ionic,
or hydrogen bonds across the interface (Figure 5.2). Although chemical bonding
is perhaps the most understood theory in adhesion principles, in many cases
chemical bonds cannot form between the components due to their dissimilar
chemical moieties, therefore, the use of coupling agents must be employed.
Coupling agents contain chemical functional groups that will coreact with the
surface of the ﬁller as well as with the matrix. Theoretically, if this occurs, the
coupling agent can act as a vehicle to bond the matrix and ﬁller with a chain of

primary bonds and, in theory, produce the strongest type of interfacial bonding.30

86

///////////J

‘5“...

B ...M
L
/////////////

 

Figure 5.2: Schematic diagram of chemical bonding between interfaces

Mechanical Interlocking
The oldest theory of adhesion is the mechanical interlocking theory, which

states that good adhesion occurs only when an adhesive penetrates into the
pores, holes, crevices, and other irregularities of the adhered surface of a
substrate, and locks mechanically to the substrate (as shown in the schematic in
Figure 5.3). However, this method has not been considered to be a viable

mechanism of adhesion in polymer composites. "1

/

l ' '

/,

Figure 5.3: Schematic diagram of mechanical interlocking

Interdiffusion
The interdiffusion theory involves the diffusion across the bond interface to

produce entanglement of polymer chain molecules. It is believed that the
molecular chains interpenetrate at the interface where the adhesive and surfaces
being bonded meet as shown in Figure 5.4. In essence, the two surfaces become

interlocked at a molecular level and, therefore, become one. This theory pertains

87

exclusively to the bonding of polymers and has relative applicability composite

applications.

 

 

1.

15
981

Figure 5.4: Schematic diagram of interdiffusion adhesion mechanism

00

 

 

 

 

p------ --—--1

 

 

EXPERIMENTAL

Materials

PLA (30510) was obtained from NatureWorks. Talc (Artic Mist) was
supplied by Luzenac America, Inc., having a median diameter of 2 microns. Prior
to extruding, the PLA and talc were separately dried overnight in an oven at 80°C
to remove excess moisture. MAPLA (containing 0.7% grafted MA), which was

4,3233 was

previously obtained and characterized by methods outlined in Chapter

also dried for 24 hours and used.

Methods

Compounding PLA/Talc Composites
PLA/talc composites were compounded using a Century ZSK-30 twin-

screw co-rotating extruder. PLA pellets were pre-mixed with 40% talc and varying
amounts of MAPLA. The MAPLA was incorporated by replacing 3%, 5%, and

10% of the PLA matrix. After mixing, the components were fed into the extruder

88

at 100 rpm and extruded under the conditions listed in Table 5.1. The extruded
strands were collected, pelletized, and dried at 80°C for 24 hours before further

processing.

Table 5.1: Extrusion Conditions for Compounding PLA/Talc Composites

 

 

Zone 1 2 3 4 5 6 7 8 9 Die
Temggf‘tu'" 25 95 145 165 175 180 180 180 175 170

 

Motor speed (rpm): 100
Feeder speed (rpm): 150

Injection Molding

The dried extruded pellets were injection molded into tensile bars using a
Cincinnati Milacron VSX85 model equipped with an ASTM type IV mold (16.51 x
1.191 x 0.318 cm). The samples were processed under the conditions shown in

Table 5.2.

Table 5.2: Injection Molding Temperatures for PLA and PLA/Talc Composites

 

Temperature (°C)

 

 

Sample Zone 1 Zone 2 Zone 3 Nozzle Mold
PLA 210 202 199 196 24
MAPLA 182 1 77 1 77 1 71 24
PLA/Talc (60:40) 212 210 210 205 24
PLA/MAPLA/Talc
(57:3:40) 212 210 210 205 24
PLA/MAPLA/Talc
(55:5:40) 212 210 210 205 24
PLA/MAPLATaIc
(50:10:40) 212 210 210 205 24

 

89

Differential Scanning Calorimetry
The melting and crystallization behavior of the PLA-talc composites were

investigated using differential scanning calorimetry (DSC). The samples were
heated to 200°C at 20°C/min and held for 5 minutes to erase any thermal history.
The samples were then cooled to -30°C at a rate of -3°C/min to observe
crystallization behavior, then reheated to 220°C at 3°C/min. Crystallization and
melting temperatures were obtained from the cooling and second heating scans,
respectively. The degree of crystallinity of each composite was calculated
according to Equation 5.1:

AH,"

= x100% (5.1)
AH; ><¢PLA

 

Zc

where AH," is the enthalpy of fusion (in J/g) and AH}; is the enthalpy of fusion

of a PLA crystal of inﬁnite size, whose reported value is 93.7 J/g. «5PM is the

weight fraction of PLA within the sample, accounting for the talc and MAPLA
addition.
Mechanical Properties

A United Testing System (UTS) Mechanical Testing Apparatus (Model
SFM-20) ﬁtted with a 1000 lb load cell was used to determine tensile properties
of the samples. Prior to testing, the tensile bars were conditioned at 25°C in 50%
relative humidity for at least 40 hours per the ASTM D638 standard.” The
crosshead speed was set at 0.5 in/min. At least ﬁve specimens were tested and
the average values were reported. The Izod impact strength was measured using
a Testing Machines Inc. (TMI) 43-02-01 monitor/impact machine. Rectangular

specimens cut from Type I tensile bars were notched and conditioned

90

appropriately according to the ASTM 0256 standard.35 Reported values were
taken as the average five specimen measurements.
Thermo-mechanical Properties

Dynamic mechanical analysis (DMA) was performed to investigate the
storage modulus, loss modulus, and tan delta values as a function of
temperature for the PLA/talc composites. Measurements were run on rectangular
bars with dimensions of 57 x 3 x 12.7 mm using a TA 2980 DMA equipped with a
dual-cantilever bending clamp at a frequency of 1 Hz and a heating rate of
10°C/min over a temperature range of 24°C to 140°C. All measurements were
performed and properties were reported in accordance with ASTM D 4065.36

Heat distortion temperatures (HDT) of the composites were also
determined using DMA according to ASTM 064837. Using a three-point bending
clamp on rectangular specimens with dimensions of 12 x 12 x 3 mm, each
specimen was heated from 25°C to 150°C at a rate of 2°C/min. The temperature
at which the test bar was deﬂected by 0.10 inches was recorded as the HDT.
Scanning Electron Microscopy

Examination of the composite microstructure was carried out using two
separate scanning electron microscopy (SEM) techniques. For the ﬁrst
technique, injection molded specimens were cut into rectangular sections,
mounted in epoxy, and polished. Each specimen was coated with osmium then
imaged using the backscattered electron imaging mode at an accelerating

voltage of 12 kV and a working distance of 15 mm.

91

The second technique was performed on the tensile fracture surfaces. The
samples were dried overnight to remove any sorbed water before being mounted
onto an aluminum stub and sputter-coated with a thin layer (21 nm) of gold.
Micrographs were obtained using secondary electrons at an accelerating voltage
of 15 kV and a working distance of 15 mm.

Statistical Analysis

Statistical analysis was performed using SPSS for Windows. Data were

compared by one-way analysis of variance (ANOVA) and a Tukey post- ANOVA

test. The mean values were considered signiﬁcant at the 0.05 level (p s 0.05).

RESULTS AND DISCUSSION

Proposed Reaction Mechanism

It has been proposed that the maleated PLA will react with the talc through
an esteriﬁcation reaction. In Figure 5.5, this reaction is shown to occur between
the anhydride group on the functionalized PLA and the hydroxyl groups on the
edge surface of the talc particle resulting in the formation of a monoester on the

surface of the talc particle.

i“ :5? ”if \ 8..

Talc particle M APLA PLA 0 0
0
CH3
HO 0
o ‘1

Figure 5.5: Proposed reaction between PLA, talc, and MAPLA yielding a
monoester on the surface of the talc particle

92

The. molecular weight of a polymer is known to signiﬁcantly affect its
surface tension. In general, surface tension decreases as molecular weight
decreases.38 Presumably, the lower molecular weight of MAPLA would enable
good wetting during processing, allowing MAPLA polymers/oligomers to coalesce
at the interface of the PLA and talc, creating an interphase. This thin region is
believed to be responsible for stronger interaction between the phases, as
evidenced by the improvements properties that will be discussed in the

subsequent sections.

Differential Scanning Calorimetry

Table 5.3 summarizes the impact that the addition of talc and MAPLA has
on the glass transition, melting, and crystallization temperatures, as well as the
crystallinity of PLA. The talc, alone, increased the crystallinity of the PLA in the
system from 16% to 26%. This increase in crystalline fraction is in agreement
with the well known nucleating effect of mineral particles in semicrystalline
materials”:40 In investigating the efﬁcacy of an interfacial modiﬁcation induced
by a coupling agent, it is important to examine the effect that it has on the glass
transition temperature. The incorporation of MAPLA into the composite resulted
in no signiﬁcant changes in the Tg nor did it have any major effect on the melting
temperatures. (Figure 5.6). However, the addition of both talc and MAPLA did
show an increase in crystallinity (p s 0.05) when compared to the unﬁlled
polymer. In the composites, the higher crystallization temperatures (measured
upon cooling) suggests the ability of the polymer to crystallize sooner and to a

greater extent as shown in Figure 5.7. According to Feng‘", in order to fully

93

understand the effect of a maleated coupling agent on the melting and
crystallization behavior of a composite, two factors must be considered. The ﬁrst
is that MAPLA is of lower molecular weight and may have a lubricating effect on
the PLA matrix. The second consideration is of the covalent and acid-base
interactions induced between the anhydride on the MAPLA and the hydroxyl
groups on the talc. Good interaction between the two functional groups could
possibly restrict the mobility of the molecules, thus, the grafted MAPLA could
disrupt the regularity of the PLA chain and hinder crystallization. Alternatively, the
low molecular weight of MAPLA could promote chain movement and favor
crystallization. In this case the latter is believed to have occurred. The crystalline
fraction increased progressively with the level of MAPLA. The MAPLA
contributed to more freely moving polymer chains and the enhanced chain
mobility permitting higher molecular orientation and better ﬁller dispersion, which

in turn led to increased crystallization activity.42

94

Table 5.3: Melting, Crystallization, and Glass Transition Temperatures for

 

 

PLA/Talc Composites
Sample Tg (°C) Tc (°C) Tm,1 (°C) Tm,2 (°C) X(%)
PLA a a a a a
58.1110 106.9131 146.5115 153.7112 16.1111
MAPLA 55.7104" 115.6126" 144.9119" 154.1109" 25.7120"
"2:333" 56.710.7"" 101.8102" 147.71078 154.6109" 25.5111"
PLA/MAPLA/ 56.310.3"" 103.8106a 147.0102"I 1538116" 27.3110"
Talc (5723140)
T:t~“€;?&) 54.6106" 107.4110" 147.8109a 154.7104" 30010.6"
PWMAPLN 53610.5" 109.9115" 148.0113" 154.9105" 32.9102d

Talc (50:10:40)

 

Different letters within the same column represent signiﬁcant differences at the 95%
conﬁdence level.

 

 

Heat Flow (W/g)

 

-0.05 -

.<'>
N

 

——-' No MAPLA

 

3% MAPLA

—----- 5% MAPLA
—--—. 10% MAPLA

 

 

-0.25
140

145

150

Temperatu re (°C)

155

 

160

 

 

Figure 5.6: DSC thennograms of PLA/talc composites revealing the effect of
MAPLA on the melting behavior

95

 

 

0.43

 

 

 

 

 

 

 

 

-——-No MAPLA
3% MAPLA ,1
0-33 ‘ ------- 5% MAPLA 1 '1
"""100/0MAPLA ] '\
0.33 - g =,
A “‘1! "l
9 0.28 - i '-
5 0.23 "1
E
8 0.18
I
0.13
0.08 . ﬁ
85 95 105 115
Temperature (°C)

 

Figure 5.7: DSC thennograms of PLA/talc composites revealing the effect of
MAPLA on the crystallization temperature

Mechanical Properties

As was evidenced by an increase in crystallinity of PLA in the composite,
the incorporation of talc inﬂuences gradual structural changes in the system. in
this way, the interactivity between composite components, combined with the
addition of MAPLA and factors such as orientation and dispersion of talc, can
affect the properties of the composites. Thus, the mechanical behavior of the
composites was investigated to study the extent of interaction between PLA and
talc with varying amounts of MAPLA. The effects of the interfacial phenomena on
various parameters, namely, the tensile strength, elongation at break, and

Young’s modulus are presented in Table 5.4.

96

Generally, the incorporation of a ﬁller into a matrix may increase or
decrease the tensile strength of the resulting composite, depending on the ﬁller
shape. Particulate ﬁllers, such as talc tend to decrease the strength of the
composites due to their lower ability to support stresses transferred from the
polymer matrix. Interestingly, in this case the sole addition of talc had no
signiﬁcant effect on the tensile strength. Adding MAPLA, however, increased the
tensile strength from 60 MPa for the unmodiﬁed composite to 66 MPa at 3%
MAPLA up to 72 MPa at 5% MAPLA, at which the tensile strength reached a
maximum; beyond which it dramatically dropped to 24 MPa. From this, it may be
inferred that increasing the amount of anhydride groups in the system increases
the number of the acid-base interactions and the degree of hydrogen bonding,
which leads to a stronger interaction between the ﬁller and the matrix. With
excess amounts of anhydride groups, however, the coupling effects of MAPLA
could have been counteracted by hydrolysis of the bulk polymer caused by
theincreased amount of end groups which could have contributed to the decline
in tensile strength.

In all of the composites, the most prominent physical effect of the talc was
observed to be the stiffening effect. This is because the Young’s modulus of talc
(170 GPa) 43 is signiﬁcantly higher than that of PLA and also because of the
enhanced crystallization of the composites. The Young’s modulus values
increased up to an MAPLA content of 3% and remained fairly constant thereafter.
This increase is most likely associated with talc particles being more dispersed

throughout the PLA in the samples with compatibilizer. Increased stiffness could

97

also be an indication of efﬁcient stress transfer from polymer matrix to ﬁller as a
result of increased interaction.14 The tensile behavior can be correlated with the
thermal results in that the degree of crystallinity can, in general, have an
appreciable effect on the mechanical properties as it affects the extent of
intermolecular secondary bonding. While the crystalline content, due to the
presence of talc, caused an increase in the strength and modulus of the

composites, it also made it brittle, accounting for the drop in elongation.

Table 5.4: Tensile Properties of PLA/Talc Composites

 

 

Tensile Young's Elongation
Sample Strength Modulus at break
(MPa) (GPa) (%)
PLA 58.7 1 2.28‘ 3.6 1 0.2a 4.3 1 1.0a
MAPLA 60.2 1 2.16‘ 4.0 1 0.23 3.9 1 0.88‘

PLA/Talc (60:40) 59.6 1 2.2"I 7.0 1 0.4" 1.1 1 0.2"

PLA/MAPLA/Talc b C 0
(573.40) 66.0 1 3.1 10.1 i 0.1 0.9 :t 0.1
PLA/MAPLA/Talc c c c
(55:5:40) 72.4 11.8 11.1 11.9 1.5 :1: 0.1
PLA/MAPLA/Talc

d C d
(55:10:40) 24.1 i‘ 2.3 11.4 i 1.7 0.3 :l: 0.1

Different letters within the same column represent signiﬁcantdifferences at the
95% conﬁdence level.

 

Moreover, the tensile behavior of the composites suggests that the
MAPLA sufﬁciently improved the wettability of the ﬁller with the polymer, which
led to improved adhesion at the interface. In addition, the results show that only a
small amount of MAPLA is needed to act as an interfacial modiﬁer to enhance
the properties, as was seen by the sudden decrease in tensile strength with
MAPLA content. This conﬁrms the existence of a critical amount of MAPLA,

98

which has an ultimate effect on the macroscopic properties of the composite. In
this work, the goal of the interfacial agent was to bridge the surfaces of talc and
PLA by providing sufﬁcient interfacial adhesion. However, too much interfacial
agent was seen to have an adverse reaction, lending to the formation of an
interphase that is too thick, limiting interfacial interactions. Therefore, it was
postulated that there exists an optimum saturation level of MAPLA that provides
the most favorable interactions across the interface of the PLA and talc?"44

Figure 5.8 illustrates the effect of the addition of MAPLA on the notched
Izod impact strengths of the PLA/talc composites. Different factors contribute to
differences in the impact strength of the composites, such as particle size, ﬁller
rigidity and its concentration, aspect ratio of ﬁller, interaction between
components and matrix phase, nucleation, orientation, and consequent structural
changes in the matrix.45 The highest impact strengths were seen for the neat
polymer and for the composite without MAPLA whose impact strengths were
virtually identical (p 2 0.05). These results are notwithstanding reports indicating
that the impact strengths of particulate-ﬁlled brittle polymers are typically lower
than the neat matrix.46

Owing to the proposed chemical reaction between MA and talc and the
resulting adhesion at their interface, higher values of notched impact strength of
composites with MAPLA were expected. However, the impact strength was lower
for all of the specimens containing MAPLA, with no signiﬁcant differences among

them (p 2 0.05). Conﬂicting data have been reported as to whether increased

adhesion improves or worsens the material’s resistance to impact.”48 In the

99

case of this work, a plausible explanation for the decrease associated with the
addition of compatibilizer is that the MAPLA induced a stronger interfacial bond
which prevented adequate crack diverting. Because of this, the composite may
not have been allowed to absorb a substantial amount of energy and caused the

composite to exhibit brittle behavior.

I'lll

 

35

 

Impact Strength (J/m 2)
>-' '—' N N DJ
0 kl] O 'J'l O

U!
I

 

 

O

 

PLA/Talc PLA/MA- PLA/MA- PLA/MA-
(60: 40) PLA/Talc PLA/Talc PLA/Talc
(57:3:40) (55:5:40) (50:10:40)

 

 

Figure 5.8: Impact properties of PLA/talc composites
Thermomechanica/ Properties

The heat distortion temperature (HDT) or softening temperature is the
maximum temperature at which a material can support a sufﬁcient load. A
semicrystalline polymer’s HDT is very sensitive to the physical properties of a
polymer, the amount of ﬁller used, and the type of filler. For example, the
crystallinity of the polymer and the glass transition temperature play a very
important role in the heat distortion temperature. Typically, the use of inorganic
ﬁllers, especially those possessing a platelet structure, improve the heat
resistance of polymers because the ﬁller can prevent deformation of the molded

composite while promoting crystallization of the polymer near the surface of the

100

ﬁller.4 The addition of talc and MAPLA increased the crystallinity of PLA, thus,
was expected to result in an increase in the HDT of PLA. However, neither the
incorporation of talc nor the addition of 3% or 5% MAPLA had any impact on the
HDT as shown in Table 5.5. Nevertheless, when 10% MAPLA was added the

HDT dropped to 50°C from 58°C, that which was seen for the
uncompatibilizedcomposite. Since a material’s HDT is directly related to its T9,
the decrease in the HDT of the composites containing 10% MAPLA corresponds
to the decrease in T9 that was observed from the DSC data. A reasonable

explanation for these results is that when there is better adhesion at the interface
there is better wettability of the matrix, binding it to the talc platelets. This, in turn,
means higher load transfer efﬁciency between the ﬁller and the matrix, making it

more difﬁcult to withstand distortion upon heating.

Table 5.5: Heat Distortion Temperatures of PLA/Talc Composites

 

 

Sample name HDT (°C)

Neat PLA 56812.0"
PLA/Talc (60:40) 58.1103"
PLA/MAPLA/Talc (57:3:40) 55410.3"
a

PLA/MAPLA/Talc (55:5:40) 55010.0
b
PLA/MAPLA/Talc (50:10:40) 50.4:0.1

Different letters represent signiﬁcant differences at the 95% confidence level.

 

To further investigate the effect that MAPLA has on the ﬁller-matrix
interactions, DMA was also used to study the response of the composite to

periodic forces. Such experiments enabled the mechanical damping and elastic

101

modulus to be measured simultaneously.49 In doing so, it was possible to obtain
valuable information on the relaxation mechanisms correlated to the
microstructure of a material. The peaks shown in the loss modulus (G”) and loss
factor (tan 6) curves are related to the molecular motion of the polymer50 and the
storage modulus (6’) curve provides information on the elastic behavior.

Figure 5.9 shows representative curves of the storage moduli (E’) of PLA
and its composites as a function of temperature. Consideration over the entire
temperature span shows that all of the specimens followed an identical trend.
The curves of the composites and the PLA each showed signiﬁcant decreases in
modulus in the range of 65-75°C. This temperature range correlates to the region
where the molecules undenNent a transition from being glassy to rubbery and
there was minimal contribution of stiffness to the matrix.51 At temperatures below
this transition the incorporation of talc in the PLA matrix increased the elastic
component of the system. This was expected since most inorganic ﬁllers,
including talc, have moduli at least one order of magnitude higher than most
polymers. The progressive increase of E’ within the low-temperature region with
MAPLA is indicative of the reinforcement effect, which allows effective stress
transfer from the PLA matrix to the talc ﬁller, leading to a higher number of PLA-
talc interactions and an increase in the load-bearing capabilities52 Consideration
of E’ at the glass transition temperatures (the respective inﬂection points of the
curves) gives insight into the relationship between the storage modulus and the
compatibilization effect. The highest modulus value corresponds to the 5%

MAPLA system, followed by 10% MAPLA, 3% MAPLA, 0% MAPLA, and lastly

102

PLA. The increase in storage modulus can be attributed to enhanced interfacial
adhesion. Beyond the glass transition region another plateau region is apparent.
This recovery of E’ is an effect typical of cold crystallization of the amorphous

regions in the polymer, a phenomenon commonly seen in PLA.53

 

 

 

 

 

 

 

 

 

 

30
_________ _.._.. PLA

25 ‘ ‘1 —--'-NoMAPLA
a l ----5% MAPLA
c1 20 ___ ~ I, 10%MAPLA
£3, ~ ............. 1i ~~---- 3% MAPLA
a .
3 15 - l
U .
O
E
a: 10 1
O)
E
.9
a) 5 1

0

35 55 75 95 115 135
Temperature (°C)

Figure 5.9: Storage modulus curves of PLA and PLA/talc composites

Figure 5.10 shows curves representing the tan 6 trends as functions of
temperature for PLA and its talc-ﬁlled composites. Tan 6 is deﬁned as the ratio of
the loss modulus to the storage modulus (E"/E’). It is sensitive to the balance of
heat dissipated and energy stored and can be used to detect thermochemical
relaxations. The tan 6 peaks remained at nearly the same temperatures for all of
the composites except for that containing 10% MAPLA, which shifted to a lower

temperature. This correlates with the previously discussed concept of the

103

plasticization effect of high MAPLA concentrations. Peak intensities were
affected by both the presence of talc and the compatibilization effect. Talc
reduced the mobility of the polymer chains, resulting in lower tan 6 values. The

MAPLA addition also decreased the tan 6 values accordingly.

 

3
_.._. PLA
2.5 _ ’ -----NoMAPLA
,: '-, ----— 3% MAPLA
_; I —---5% MAPLA
2 ' ‘ ”1,. —-10%MAPLA

 

 

 

 

 

115

Temperature (°C)

Figure 5.10: Tan 6 curves of PLA and PLA/talc composites

For the compatibilized composites, it is possible to observe a correlation
between the intensity of the tan 6 peak and the interfacial adhesion. All of the
curves displayed a similar tendency until a temperature of about 80°C was
reached, at which the 10% MAPLA curve showed a small shoulder. This effect
could imply that at 10% of MAPLA the existence of a third component, rather

than an interfacial additive in a two-phase system.39

104

The loss modulus curves, E”, are shown in Figure 5.11. The temperatures
denoting the maxima of the loss modulus curves are representative of the glass

transition temperatures due to the motions associated with unrestricted

amorphous PLA. From the ﬁgure it can be seen that the T9 value for the

composite containing 3% MAPLA shifted to a slightly higher temperature, the T9

of the 10% MAPLA system shifted to a lower temperature, while the 5% MAPLA
system remained the same as the neat polymer. This suggests that 5% MAPLA

may be a more efﬁcient compatibilizer than the other concentrations. It should be

noted, however, that the Tgs obtained from DMA are quite different from those

obtained from the DSC thennograms and may even show a slight contradiction.
This discrepancy can be attributed to a difference in chain mobility as a result of
different experimental conditions. The kinetics of the glass transition, the heating
rate, and the loading frequency each has some impact on the molecular mobility.
Therefore, DMA is a more sensitive tool than DSC In determining the thermal
transitions of polymers.54

Furthermore, the peak heights of the loss modulus curves are associated
with the damping of the material and are related to the energy converted into
heat upon deformation. The composites had higher loss moduli than PLA
because the presence of talc disrupts the polymer bulk, as the mineral particles
are coated by the amorphous phase. The addition of MAPLA at the interface
further increased the loss modulus values. The highest peak corresponds to the

5% MAPLA composite, followed by the 10% MAPLA composite, the 3% MAPLA

105

composite, and ﬁnally the unmodiﬁed composite. Furthermore, a pronounced
broadening was seen in the loss moduli of the compatibilized composites,
bringing about a transition region with high damping over a temperature range a
few times greater than that of the unﬁlled polymer. The broadening effect as well
as the increase in the E” can be attributed to a combination of the high surface
area of the ﬁllers and the strong interaction between PLA and talc. It is possible
that the stronger interaction resulted in the partial immobilization of the polymer
chains due to adsorption of polymer on the talc surface. The restriction of the
polymer chains induced by talc caused a change in density of the packing of the

. polymer and increased the loss modulus.

 

 

7
-...... PLA
6 . ii —---- No MAPLA
l 1‘ ----- 3% MAPLA
5 - 1‘ —-— 5% MAPLA
3 — 10% MAPLA
4 « l

Loss Modulus (GPa)
(a)

 

 

 

 

45 65 85 105 125
Temperature (°C)

Figure 5.11: Loss modulus curves of PLA and PLA/talc composites

106

Scanning Electron Microscopy

Composite Microstructure
The morphology and microstructure of composite materials play a

signiﬁcant role in the prediction of macroscopic properties. The composite
microstructure was examined using backscattered electrons to effectively
observe the size, dispersion, and orientation of the talc particles within the PLA
matrix. Unlike secondary electrons, backscattered electrons are primary beam
electrons that have been elastically scattered by nuclei and are strongly
dependent on the average atomic number of the sample. The backscattered
electron images of the polished samples clearly show contrast based on the local
chemical composition of the material. Talc, having a higher average atomic
number than PLA, is revealed as white particles in the micrographs denoted by
the arrows in Figures 5.12 through 5.15. In the sample that contains no MAPLA
(Figure 5.12) large talc agglomerates are present. The talc particles are relatively
unevenly distributed and show no orientation. This conﬁguration is as opposed to
that seen in the samples that contain MAPLA (Figures 5.13 through 5.15), all of
which display talc particles that are more evenly dispersed within the matrix and

exhibit axial orientation.

107

 

 

Figure 5.12: Backscattered electron image of PLA-talc composite containing no

108

 

 

Figure 5.13: Backscattered electron image of PLA-talc composite containing 3%
MAPLA

109

 

Figure 5.14: Backscattered electron image of PLA-talc composite containing 5%
MAPLA

110

 

 

Figure 5.15: Backscattered electron image of PLA-talc composite containing 10%
Fracture surface analysis

Postmortem investigation of the PLA-talc composite fracture surfaces
provided signiﬁcant insight into fracture mechanisms and morphological changes
that led to changes in fracture behavior. The microstructures of the composites
were correlated with their fracture surfaces to help elucidate the mode of failure.
Each of the composites’ topographical micrographs was examined to lend
evidence to the assertion of improved interfacial adhesion with added MAPLA.
Although all of the composites demonstrated brittle behavior .due to the inherent
brittle nature of PLA, one of two failure mechanisms was shown to be

dominant—(1) failure of the PLA matrix or (2) debonding via platelet pull-out.

111

The micrograph of the composite without added MAPLA displays typical
characteristics of an incompatible composite (Figure 5.16). In it, a large number
of voids appear resulting from talc platelet debonding. The distinct edges of the
talc particles with no polymer adhered to the surface indicate behavior
representative of a muIti-component system in which there is little or no adhesion
at the interface.

In the composite containing 3% MAPLA, shown in Figure 5.17, it can be
seen that many of the talc particles are imbedded within the polymer matrix. Less
distinction is seen between talc and PLA. Fewer cavities and edges are present
around the talc particles and the talc platelets are coated with PLA, which
indicates an increase in adhesion. In addition, the talc platelets are more
imbedded within the polymer, which causes the talc particles to appear smaller
than in the uncompatibilized composite.

In Figure 5.18, the fracture surface of the composite with 5% MAPLA
shows a very different morphology from the other composites, one that is quite
distinctive of a ductile failure. The morphology of this specimen shows signiﬁcant
localized deformation of the PLA matrix in the form of ﬁbrils. It is also evident that
the talc particles are more deeply lodged within the matrix and that failure occurs
within the polymer, as shown by the defamation of the PLA rather than by
debonding at the interface. This suggests better wetting of the talc by the PLA in
this system. These observations substantiate the higher mechanical properties
that were exhibited in this sample, since stronger adhesion permits an applied

load to be effectively transferred from the matrix to ﬁller.

112

The 10% MAPLA system, in Figure 5.19, shows behavior characteristic of
a brittle material. In materials that undergo brittle fracture there is rapid crack
propagation and very little plastic defamation, yielding a ﬂat surface. It was also
interesting that throughout this particular sample composition, two
microstructures were apparent: a very ﬂat surface and one in which multiple
voids were present. This is likely caused by the high concentration of MAPLA
exceeding its solubility limit in PLA, resulting in a heterogeneous microstructure
and acting as a detriment to the mechanical behavior. It is necessary to note that
it was this composition which displayed the poorest tensile strength and. the
lowest elongation at break. The brittle behavior exhibited by this composite is
most likely due to either excess interfacial adhesion between talc and PLA,
preventing safe debonding of the constituent phases, or to increased degradation
of the composite clue to excess MAPLA which lowered the molecular weight of
the polymer. These electron micrographs further coninn the existence of a critical
value of MAPLA which is necessary to improve the adhesion in PLA/talc

composites.

113

PLA matrix:

‘,../'/"‘
. I" . ‘1'
. "1"..ﬁ” .’
. - 1’
‘ ”I ' ’4
I..~' ' /
1."..- II
I ..
1"
I“. ‘

,1.
I!”

I

M,

'h

v

. Talc platelet

 

Figure 5.16: Secondary electron SEM micrograph of PLA-talc composite with no
added MAPLA indicating little or no adhesion between talc and PLA as
evidenced by platelet pull-out

114

’0

PLA matrix
4r.-
.x"’\.,\

A

i
.v _ ".1:
- * it»;
Talc particles 3 1
'\ 4
\

 

Figure 5.17: Secondary electron SEM micrograph of PLA-talc composite
containing 3% MAPLA denoting slight to moderate adhesion between talc and
PLA

115

g -"7‘-—«.‘a

i-
» :-

PLA matrix 1,13,.

 

Figure 5.18: Secondary electron SEM micrograph of PLA-talc composite
containing 5% MAPLA revealing enhanced adhesion and a more ductile fracture

116

0:“

Ic. ,

‘ .\ platelet _ f
' . _' ' ‘1 .2 (

PLA matrix . \ "Ta

 

Figure 5.19: Secondary electron SEM micrograph of PLA-talc composite
containing 10% MAPLA displaying a ﬂat surface texture, indicating brittle
composite failure in the PLA matrix

CONCLUSIONS

MAPLA was used as an interfacial modiﬁer in PLA-talc composites.
Investigation of the thermal, mechanical, and morphological properties provided
experimental evidence of increased adhesion at the interface. These data
indicate the existence of a critical value, beyond which no more improvements
occurred as reﬂected by a sudden decrease in tensile strength. Microstructural
observation of the composites revealed the modes of failure and the dispersion
of the talc particles with the matrix, both of which were in accordance with the

mechanical properties analyses.

117

REFERENCES

1. Hiljanen-Vainio, M., Heino, M., and Seppéla, J.V., Polymer, 1998, 39, 865.
Hu, R., Lim, J.K., J. Comp.Mater., 2007, 41, 13, 1655.

Oksman, K., Skrifvars, M., Selin, J.-F., Comp. Sci. Tech., 2003, 63, 1317.
Serizawa, S., lnoue, K., Iji, M., J. Appl. Polym. Sci., 2006, 100, 618.

Huda, M., Drzal, L., Misra, M., Mohanty, A., J. Appl. Polym. Sci., 2006,102,
4856.

Ke, T., Sun, X., J Appl Polym Sci, 2003, 89, 1203.

7. Ogata, N., Jimenez, G., Kawai, H., Ogihara, T., J. Polym.Sci.: Part B: Poly
Physics, 1997, 35,389.

8. Wan, Y.Z., Wang, Y. L., Xu, X. H., Li, O. Y., J. Appl. Polym. Sci., 2001, 82,
1, 150.

9. Zimmerman, M. 0., Alexander, H., Parsons J. R., Bajpai, P. K. ACS
Symposium, 1991, 457 132.

10.Mohamed, A., Gordon, S., Carriere, C., Kim, S., J. Food Quality, 2006, 29,
266

11.Urayama, H., Ma, 0., Kimura, Y., Macromol. Mater. Eng., 2003, 288,562.

12.Thakur, K. A.M., Kean, R.T., Zupfer, J.M., Buehler, N. U., Doscotch, M.,
Munson E., Macromolecules, 1996, 29,27, 8844.

13.Kolstad, J., J. Appl. Polym. Sci., 1996, 62, 1079.

14.Huda, M., Drzal, L., Misra, M., Ind. Eng. Chem. Res, 2005, 44, 5593.
15. Klyosov, A.A., Wood-Plastic Composites; Wiley, 2007.

16.Plackett, D., J. Polym. in the Environ., 2004, 12, 131.

17.Febrianto, F., Yoshioka, M., Nagai, Y., Tahir, P., Syaﬁi, W., Shiraishi,
N.,J.Bio. Sci, 2006, 6, 555.

18.Petersson, L., Oksman, A. Mathew, P., J. App. Polym. Sci, 2006, 102, 1852.

19.Michot, L. J., Villieras, F., Francois, M. Yvon, J., Le Dred, R., Cases, J. M.,
Langmuir, ACS, 1994,10, 3765.

20.Kaw, A. K., Mechanics of Composite Materials; CRC Press, 2006.

@9990)

9’

118

21.Strong, B., Fundamentals of Composites Manufacturing: Materials, Methods
and Applications; SME, 2007.

22.Jones, R. M., Mechanics of Composite Materials; Taylor & Francis, 1998.
23. http://www.azom.com.

24.Liebennann, H.H., Rapidly Solidiﬁed Alloys: Processes, Structures,
Properties, Applications; CRC Press, 1993.

25.Peter Kollar, L.P., Springer, G.S., Mechanics of Composite Structures;
Cambridge University Press, 2003.

26.George,J., Sreekala, M. S., Thomas, S., Polym. Eng. Sci., 2001,
41,9,1471.

27. Fisher, G., MS. Thesis, Michigan State University, East Lansing, MI, 1994.

28.Yosomiya, R., Morimoto, K., Nakajima, A., Adhesion and Bonding in
Composites; CRC Press, 1990.

29.Toshev, B. V., Int. J. Mech. Eng. Ed. 2006, 34, 3, 256.

30. Erickson, P. W.; Plueddemann, E. P., Historical Background of the
Interface—Studies and Theories. In Interfaces in Polymer Matrix Composites,
Plueddemann, E. P., Ed.; Academic Press: New York, 1974, Chapter 1.

31.Stamm, M., Polymer Surfaces and Interfaces: Characterization, Modiﬁcation
and Applications; Springer, 2008.

32.Carlson D., Nie, L., Narayan, R., Dubois, P., J. Applied Polym. Sci., 1999,
72, 477.

33. Li, H.-M., Chen, H.-B., Shen, Z.-G., Lin, 3., Polymer, 2002, 43, 5455.
34.ASTM Standard D638.

35.ASTM Standard D256.

36.ASTM Standard D4065.

37.ASTM Standard 0648.

38.Park, H., Park, 0.3., Tzoganakis, C., Chen, P., Ind. Eng. Chem. Res, 2007,
46 (11), 3849.

39.Garcia-Martinez, J.M., Laguna, O., Areso, S., Collar, E. P., J. Appl. Polym.
Sci., 2001, 81.

40.Karrad, S., Lopez Cuesta, J.-M., Crespy, A., J. Matl. Sci., 1998, 33,2, 453.

119

41.Feng, D., Caulﬁeld, D., Sanadi, A., Polym. Compos, 2001, Vol. 22, 4, 506.

42.Jang, W.J., Shin, B. Y., Lee, T.J., Narayan, R., J. Ind. Eng. Chem, 1007, 13,
3,457.

43.Martinatti, R., Ricco, T., J. Mater. Sci., 1994, 29, 442.
44.Sun X., Zhang, J.-F., Biomacromol, 2004, 5, 1446.

45.Wypych G. Handbook of ﬁllers. 2nd ed. Toronto: ChemTec Publishing; 1999,
395—460.

46.Young, R.J., in Structural Adhesives: Developments in Resins and Primers,
Elsevier, 1986, 163.

47. Phillips, D.C., Harris, B., Polymer Engineering Composites; Applied Science:
London,1977,45.

48. Spanoudakis, L., Young, R. J., J. Mater. Sci., 1984, 19, 473.

49.Nielson, L., Mechanical Properties of Polymers, Chapman and Hall: London,
1962: 138.

50.Young, R. J., Lovell, P. A., Introduction to Polymers, 2nd ed.; Chapman and

Hall: London; New York, 1991 ; 335.
51.Rana, A. K., Mittra, B. C., Banerjee, A. N., J. App. Polym. Sci, 1999, 71, 531.

52.Shah, Bhavesh, J. Vinyl Addit. Technol, 2005, 11, 160.

53.Nazhat, S. N., Kellomaki, M., Tanner, K. E., Bonﬁeld, W., J. Biomed. Mater.
Res., 2001, 58, 335.

54.Thermal Characterization of Polymers, 2'"d ed.; Turi, E., Ed.; Academic Press:
San Diego, 1991; vol.1.

120

Chapter 6

 

REACTIVE MODIFICATION OF POLYLACTIC ACID BY SILANE GRAFTING

AND CROSSLINKING REACTIONS

 

INTRODUCTION

PLA has become a polymer of increasing commercial interest and is the
biobased material of choice for many, owing to its good physical and mechanical
properties. In spite of its many advantages, some drawbacks have hindered its
growing success in a number of industrial markets. One major limitation of PLA is
its processing instability which is generated by the thermal, oxidative, and
hydrolytic degradation that occurs during processing. This results in a breakdown
of chain integrity and a reduction in molecular weight, both of which have
detrimental effects on the rheological properties of the melt and the mechanical
properties of the product. This, together with PLA’s inherent low melt strength,

has motivated considerable research efforts to overcome these shortcomings.

121

Most of such efforts have been focused on increasing the molecular weight and
melt viscosity of PLA, many of which involved the introduction of functional
additives to link multiple chains together, but these have been performed with
little success."6 In other efforts common chain extenders, such as boric acid and
lactone compounds, have been employed. However, these have proven
ineffective in PLA-based polymers.7 Consequently, crosslinking has been
deemed a well-suited approach to increase the molecular weight of PLA and
improve upon some of the properties that continue to plague its widespread
usage.

Crosslinking is an important process in polymer technology that involves
the formation of a three-dimensional network of polymer chains through covalent
or ionic bonding. The incorporation of crosslinks within a polymer can provide
substantial improvements in its properties. Oftentimes crosslinked polymers ﬁnd
use in applications that require good thermal and dimensional stability, chemical
resistance, and electrical properties.

Figure 6.1 shows a schematic illustration of the structural differences
between a crosslinked polymer network and a linear polymer. The links between
the polymer chains in the crosslinked network add important features to the
polymer. They allow a semicrystalline polymer, such as PLA, to exhibit
mechanical properties characteristic of a thermoplastic below its melting
temperature. In addition, they permit the polymer to resist ﬂow and exhibit the
mechanical properties of a rubber when heated above its melting temperature."

With this in mind, crosslinking PLA may not only improve its high temperature

122

properties, but it might also improve its melt strength and enhance its molecular
weight to compensate for the degradation associated with thermal processing."22

This project introduces the concept of silane grafting and crosslinking to
post-polymerized PLA. A sequential series of experiments was performed to
monitor the effects of various parameters, such as concentrations of silane,
peroxide, and catalyst, with an intention to explore the structure-property
relationships related to crosslinking. The speciﬁc objectives are the following: (1)
the use of reactive extrusion to achieve radical melt functionalization of an
organofunctional silane; (2) the moisture-induced crosslinking of the silanol
moieties that are attached to the polymer chain; and (3) evaluation of the
crosslinking reaction and the characterization of the resultant crosslinked
polymers on the basis of gel content, mechanical behavior, and thermal
properties, in order to understand how the crosslinked structure affects the

properties of PLA.

 

Polymer
crosslinks

 

 

 

 

Linear
chains

 

 

 

Figure 6.1: Schematic illustration of a crosslinked and uncrosslinked polymer

In the paragraphs that follow, an introduction to polymer crosslinking

including types and mechanisms, as well as a review of past work with particular

123

regard to PLA, will be given. In addition, a brief theoretical background as it

pertains to the nature of this work will be provided.

PLA Crosslinking Background

The effects of crosslinking on the properties of PLA have been well
studied. The most reported and undoubtedly the most preferred route, according
to the literature, is associated with peroxide-induced radical reactions in which
peroxides are incorporated into PLA via extrusion processes. In this method, the
role of the peroxide is two-fold: (1) to stabilize PLA by deactivating the residual
catalysts that promote thermal degradation and (2) to introduce free radicals that
can be linked together by coupling reactions.23 This multi-step process is done
by, ﬁrst, the generation of free radicals induced by the thermal decomposition of
a peroxide. These free radicals are then capable of extracting hydrogen atoms
from the polymer backbone and transferring the free radical site to the PLA
chain, forming a PLA macroradical. Lastly, two neighboring PLA chains
containing active radicals can undergo bimolecular recombination to produce a
crosslinked PLA network.

In the crosslinking of PLA via the peroxide-generated radical technique,
one of the most important factors that must be considered is the peroxide type
and whether it is capable of abstracting hydrogen from PLA. Additionally, the
peroxide’s half-life and the extruder’s residence time play critical roles in the
degree of peroxide decomposition, rate of cure, and the ultimate extent of
crosslinking. Various peroxide types have been investigated and results have

been presented on the optimum stabilization and crosslinking effects in PLA. For

124

example, Sodergard et al.23 found that when dialkyl peroxides having no carbonyl
groups were used, such as dicumyl peroxide (DCP), the results yielded PLA with
improved elongation at break and reduced crystallization rates, yet with
increased rates of hydrolytic degradation. Nevertheless, when peroxides
containing carbonyl groups were used the opposite was seen and the hydrolytic
activity of PLA was stabilized. Nijienhuis et al.9 also noted that DCP resulted in
the degradation of PLA and that the decomposition products induced a
plasticization effect on the polymer. Other researchers, including Carlson et al.”,
have shown Luperox 101, a difunctional di-tertiary alkyl peroxide, to be quite
effective in inducing free radical branching and crosslinking of PLA. Peroxide
decomposition rates and molecular weights were studied by Takamura et al.1
who assessed various peroxides based on crosslinking efﬁcacy in terms of free
radical efﬁciency and hydrogen abstraction ability. Their results pointed toward
using peroxides with higher decomposition rates to obtain a higher degree
crosslinking. This is because the formation of more radicals leads to a higher
probability of crosslinking, allowing it to predominate over chain scission.25 Based
on these results, the main difﬁculty with using peroxide to crosslink PLA lies in
the nature of the radical reactions. Often a surplus of radicals in PLA causes it to
be more susceptible to chain scission, giving rise to a greater number of
oligomers. The increased number of hydroxyl and carboxylic end groups promote
thermal decomposition, resulting in lower molecular weight and decreased

properties. 24

125

Similar to peroxide-induced crosslinking is the irradiation of polymers to
form a network structure. This technique, ﬁrst introduced in the 19508,26 has
been proven to be effective for the grafting, crosslinking, and degradation of
many polymers, typically polyoleﬁns. However, in recent years, it has also been
utilized in polyesters.9'13'1“'27'28 In this technique, the formation of macroradicals is
based on hydrogen expulsion via irradiation by gamma or electron beams,
yielding macroradicals that can combine with adjacent chains to form three-
dimensional networks by carbon-carbon linkages.

Expensive equipment and the necessary high voltages coupled with
difﬁculty producing sections with even irradiation represent the major limitations
of this technique. Furthermore, in PLA, the direct exposure to gamma-ray or
electron beams through ionization results in degradation at low irradiation doses
and embrittlement at higher irradiation doses."29 Consequently, the use
polyfunctional monomers (PFMs) as crosslinking agents in combination with
irradiation has been investigated to combat the issues with degradation with
some success. PFMs are very reactive and their ability to react with PLA can be
utilized to form a crosslinked structure. According to Nagasawa et al.”, 3% of
triallyl isocyanurate (TAIC) is effective in producing crosslinked PLA, manifested
by a thermal stability exceeding 200°C. The increase in thermal stability was
attributed to the cyclic unit in TAIC and the large three-dimensional network that
it produces. Other researchers, including Jin et al.31, Mitomo et al.”, and Quynh

L32

et a , conﬁrmed these ﬁndings, reporting the use of TAIC to result in high gel

contents and improved thermal stability of PLA.

126

Silane Crosslinking

The crosslinking technique of interest in this work is silane crosslinking
using organic silicon compounds. Limited evidence exists in the published
literature of silane crosslinking in aliphatic polyesters and no studies have been
reported on silane-crosslinked post-polymerized PLA. In general, this method
entails grafting an organofunctional silane onto a polymer backbone and
subsequently crosslinking in a moisture-rich atmosphere.

Organofunctional silanes are bifunctional. Bound to the central silicon
atom are two different functional groups, an organofunctional group and a
silicofunctional group. The organofunctional group is strongly attached to the
silicon atom by a highly unreactive carbon chain. The silicofunctional groups-
those which are directly bound to the silicon atom, are hydrolyzable with relative
ease.33 Upon hydrolysis to silanol groups, the silicofunctional groups may react
with the hydroxyl groups of an inorganic or organic material or by condensation
to form stable siloxane linkages, which are then responsible for the formation of
the resultant network structure. The higher bond energy associated with the Si-Si
bonds induced by silane crosslinking leads to much less thermal degradation by
chain scission when compared to free-radical crosslinking (C-C) bonds.
Therefore, a major advantage of silane crosslinking is the ability to yield products
with much higher thermal stability than peroxide or irradiation crosslinked
materials.

The concept of silane crosslinking was ﬁrst introduced in the late 19603,
when Dow Corning patented a two-step crosslinking method using organic

peroxides and vinyl silanes to improve the high temperature properties, chemical

127

resistance, and mechanical properties of polyethylene, all without the high
investment costs of irradiation equipment and the risk of pre-curing.“ This
technology, based on organic silicon chemistry, was ﬁrst established in an
attempt to improve the properties of polyethylene in cable insulations and hot
water pipes, but has since been expanded to other applications. The process,
known as SioPlas technology, involves the peroxide initiated melt grafting of vinyl
silane onto a polyethylene backbone and then crosslinking in the presence of
water."5 Crosslinking will occur spontaneously at ambient conditions, however, it
is a very slow process, which is why a masterbatch of a suitable catalyst is
typically used prior to the condensation step to accelerate cure.

Another technique, similar to the SioPlas method, is the Monosil
technique, which was patented by Maillefer and BICC.3" Using this one-step
approach, polyethylene pellets, an antioxidant, and peroxide are combined into a
high-shear extruder. Next, meter-fed silane and catalyst are pumped into the melt
to ensure crosslinking during processing, eliminating the need for multiple steps.
The resultant products obtained from the Monosil process are comparable to
those produced from SioPlas technology.

With the advent of SioPlas and Monosil technologies, silane crosslinking
has been an ever-growing ﬁeld. This technology has been fully exploited in

37-45

polyethylene and has expanded to other polymers such as polyvinyl chloride

4&5", and polyethylene-octenesz. However, as was previously

46"", polypropylene
mentioned, the modiﬁcation of PLA by the grafting of an organofunctional silane

and the subsequent crosslinking has yet to be explored as an avenue for

128

increasing the molecular weight, enhancing the melt strength, and improving the

hydrolytic stability of PLA or any other aliphatic polyester.

Network Theory

The formation of a crosslinked material begins by the branching of linear
chains, followed by their linking together into a network whose molecular weight
diverges to inﬁnity. This irreversible transformation of a linear polymer into a
therrnoset material is known as gelation and the point at which it occurs is
deﬁned as the gel point. It is at this critical point beyond which the polymer loses
its ability to ﬂow. Once a polymer is completely crosslinked it is not readily
thermally processable and is insoluble.

Chemical crosslinks in a polymer at the molecular level are manifested by
macroscopic effects, such as changes in molecular weight, mechanical behavior,
crystallization kinetics, melting, and glass transition behavior, as well as
differences in thermal stability. The extent to which these properties are affected
is governed by both the properties of the constituent monomers and by the
crosslink density.” The crosslinking density is deﬁned as the number of
crosslinks in a unit volume of a polymer network and can be correlated with the

distance between crosslinking points by Equation 6.1: 9
V = 370» (6.1)

where v is the crosslink density, p is the density of the polymer, and Me is the

molecular weight between crosslinks. As a general rule, the greater the crosslink

density, the tighter the mesh size, and the more apparent are the property

129

changes. Topological constraints, such as polymer loops and entanglements,
also play a role in altering the properties, but are effective only if the minimum
crosslink density is such that it prevents the chains from uncoiling.54 The
determination of crosslink density can be achieved by a number of methods. One
method is by measurement of the swelling equilibrium. When a crosslinked
polymer is immersed in a good solvent, rather than dissolving the polymer, the
solvent molecules diffuse into the network, solvating the polymer chains. This
breaks the intermolecular forces and the network swells. An elastic retractive
force is generated by the polymer opposing deformation and establishing
equilibrium between the solvation forces that drive the chains apart and the
elastic forces of the network that hold the chains together.“57 This equilibrium
can be characterized by the steady-state swelling ratio. Given the direct
correlation with the crosslink density, swelling ratio determination is a simple and
cost-effective tool to characterize crosslinked polymer networks. The steady-
state swelling ratio can be described using Equation 6.2:

v2 = ——1— (6.2)

1+Q&
P1

where v2: V.,/V (V0 is the volume of the uncrosslinked network and V is the

volume of the crosslinked network). Q is the swelling value deﬁned as the weight

ratio of the crosslinked polymer before and after immersion in the solvent, pr and

p1 are the densities of the polymer and the solvent, respectively. The number of

130

crosslinked monomer units per main chain in the absence of solvent can then be

deﬁned mathematically by JP. Flory's equation (Equation 6.3)"":

 

v :_ 1n (1 —v2) +v2 + Zlvzz (6 3)
91 021/3— v2/2>

Where vx is the crosslink density, v2 is deﬁned above, Z1 is the Flory interaction

parameter, and CD1 is the molar volume of the solvent. Knowledge of the Flory
interaction parameter permits the crosslink density to be calculated. The

molecular weight between crosslinks, M c , can then be determined by Equation

6.4:

M. =_—-. (6.4)

where v, was previously deﬁned as the crosslink density and 17 is the speciﬁc

volume of the polymer.

EXPERIMENTAL

Materials

PLA (3051 D) was obtained from Natureworks. The silane used in this
work was Vinyltrimethoxysilane (VTMOS) (Figure 6.2) and was purchased from
Gelest. The free-radical initiator, 2, 5-dimethyl-2, 5-di-(tert-butylperoxy) hexane
(L101) (Figure 6.3) and the catalyst, dibutyltin dilaurate (DBTDL) (Figure 6.4)

were obtained from Sigma. All reagents were used as received.

131

9H2
H3CO-Si-OCH3
00H3

Figure 6.2: Structure of vinyl trimethoxysilane (VTMOS)

CH3 CH3 CH3 CH3

I I H2 H2 I I
H3c—clz—o—o—c—c —c —<I:—o—o—<I3—CH3

CH3 CH3 CH3 CH3

Figure 6.3: Structure of 2, 5—dimethyl-2, 5-di-(tert-butylperoxy) hexane (L101)

C4H9
o I 0
fO—SIn—Oﬁ
C11H23 C4H9 C111123

Figure 6.4: Structure of dibutyltin dilaurate (DBTDL)

Methods

Preparation of PLA-g-VTMOS
In the ﬁrst step, VTMOS was grafted onto PLA by a free-radical reaction.

Solutions of 0.1% L101 with varying VTMOS percentages (0.5% to 3%) and of

3% VTMOS with varying L101 (0.05% to 0.5%) were prepared. Sample

compositions are tabulated in Table 6.1. The mixtures were then manually

combined with the PLA pellets to ensure even coating before being introduced

into the feeder of a Century ZSK—30 twin-screw extruder. The polymers were

extruded according to the processing conditions in Table 6.2. The extrudate was

132

cooled through a water bath and pelletized downstream. Selected samples were
subsequently dried and combined with varying amounts of DBTDL and re-
extruded under the same extrusion conditions. Following extrusion, the samples
were dried and stored in a desiccator prior to further processing by injection

molding according to the conditions in Table 6.3.

Table 6.1: VTMOS Grafting Sample Compositions

 

without cata_ly_s_t

L101 (%) VTMOS (%)

0.1 0

0.1 0.5

0.1

0.1

0.1

0.05

0.2

0.5

 

00000001004

with cJatalvst
L101 (%) VTMOS (%) DBTDL (%)

 

0.1 0 0.05
0.1 0.5 0.05
0.1 1 0.05
0.1 3 0.05
0.1 5 0.05
0.05 3 0.05
0.2 3 0.05
0.5 3 0.05
0.2 3 0.1
0.2 3 0.2

 

Table 6.2: Extrusion Processing Conditions for VTMOS Grafting

 

 

 

Zone: 1 2 3 4 5 6 7 8 9 Die
Temperature:
(°C) 25 95 145 175 180 180 180 180 175 170
Motor speed:
(rpm) 100

 

133

Table 6.3: Injection Molding Processing Conditions for VTMOS Grafting

 

. Temperature (°C)
Zone Zone Zone
Sample 1 2 3 Nozzle Mold

 

 

VTMOS-g—PLA 210 202 199 196 24

 

Moisture Crosslinking
Moisture crosslinking was performed on the injection molded articles.

Since the rate of crosslinking depends on the rate of water diffusion into the
polymer specimens of the same shape and thickness were used. Crosslinking of
the catalyzed and uncatalyzed specimens was carried out in a moisture
saturated chamber at room temperature over a 28-day period. Crosslinking
properties were determined at varying curing time intervals of 1 day, 7 days, 14
days, 21 days, and 28 days. After crosslinking, all samples were dried and
conditioned for further characterization.
Melt Flow Index (MFI)

The melt ﬂow rate was used to investigate the degree of crosslinking of
PLA. Experiments were carried out using a RayRan Melt Flow Indexer. All tests
were run in accordance with ASTM D1238-90b. The pelletized samples were
melted at 190°C and driven through a capillary die using a 2.16 kg weight
attached to a piston. The mass (in grams) of the extrudate was measured at 30

second intervals and the MFI was calculated accordingly.

Gel Content and Degree of Swelling
The gel content of the crosslinked polymers was determined by Soxhlet

extraction using methylene chloride as the solvent. First, 2 grams of the dried

134

sample were introduced into a cellulose thimble and extracted over a period of 24
hours. After extraction, the residual polymer was dried to constant weight. The
gel content was calculated as the percentage ratio of the ﬁnal weight of the
polymer to its initial weight using Equation 3.8. Gel contents of at least three
specimens of each sample were determined and the average was reported. The
degree of crosslinking was determined by calculation of the swelling ratio, which
was evaluated by weighing approximately 1 gram of the dried crosslinked
polymer, immersing it in 100 mL of methylene chloride for 48 hours, removing
and then weighing the gel in its fully swollen state. The gel was later dried to
constant weight and the degree of swelling (DS) was calculated using equation

6.5:

 

DS = [(W, — Wg)/Wg][pC/:1:72 ] (6.5)

where Wg is the average weight of the dry gel component in the crosslinked

sample, W3 is the weight of the gel component swollen at room temperature for

48 hours in methylene chloride, and pp and PCHC12 are the densities of PLA

and methylene chloride, respectively.
Differential Scanning Calorimetry (DSC)

The melting and crystallization behavior of both the crosslinked and
uncrosslinked polymers were investigated by DSC analysis using a TA
Instruments 2920 DSC machine operated under nitrogen. The samples were
heated to 220°C at 20°C/min and held for 5 minutes to erase any thermal history.
The samples were then cooled to -30°C at a rate of -5°/min, then reheated to

135

la

alll

X-r

541

at

220°C at 5°C/min. Crystallization and melting temperatures were obtained from
the cooling and second heating scans, respectively.
Thermal Gravimetric Analysis (TGA)

A high resolution thermogravimetric analyzer TGA 2950 from TA
Instruments was used to determine the changes in thermal stability due to
crosslinking. Each sample was heated from 20°C to 500°C at 25°C/min. The
balance and the samples were purged with nitrogen gas at rates of 40 ml/min
and 60 ml/min, respectively.

X—ray Photoelectric Spectroscopy (XPS)

XPS spectra of crosslinked PLA were obtained using A Perkin Elmer Phi
5400 ESCA system with a magnesium Ka X-ray source. Samples were analyzed
at pressures between 10'9 and 10‘8 torr with a pass energy of 29.35 eV and a
take-off angle of 45°. The spot size was approximately 250 pmz. Atomic
concentrations were determined using previously determined sensitivity factors.
All peaks were referenced to the signature C1s peak for adventitious carbon at
284.6 eV.

Statistical Analysis

Statistical analysis was performed using SPSS software for Windows.
Data were compared by two-way analysis of variance (ANOVA). A Tukey post-
Hoc test was used to test multiple comparisons. For each analysis the mean

difference was found signiﬁcant at the 0.05 conﬁdence level (p s 0.05).

136

RESULTS AND DISCUSSION

Proposed Reaction Mechanism

A schematic illustration detailing the proposed mechanism for silane
grafting is illustrated in Figure 6.5. A two-step reaction is necessary to attain the
desired functionalized product. It is expected that L101 undergoes thennally-
induced decomposition in the ﬁrst step, forming peroxide radicals under the
extrusion conditions. The peroxide radical is then able to react with the PLA
chain and undergo radical transference to form a PLA macroradical. This PLA
macroradical breaks the silane unsaturation, enabling grafting onto the PLA
backbone. This reaction is followed by another radical transfer forming a VTMOS
grafted macroradical, which can react with other molecules of PLA to continue
the reaction until the chain terminates by chain transfer to polymer. The resultant
grafted polymer contains alkoxyl moieties that can easily be hydrolyzed to silanol
groups, which are capable of reacting with polymers or ﬁllers containing hydroxyl
groups or by condensation to form a crosslinked network. The latter, as shown in
Figure 6.6, will be the focus in the context of this work. The moisture crosslinking
reaction of the grafted silane is comprised of, first hydrolysis of the alkoxysilanes
to silanol groups, which when combined can condense to siloxane groups. Thus,
what follows is the formation of a network of PLA chains tied together by

stabilized siloxane linkages.

137

 

 

 

R-O-O-R . .. > 2 'OR
L101 peroxrde decomposrtron

 

 

(’00'139\ 0 CH3
* 'OR >
initiation . 0% + ROH
0 CH3 if: 0 CH2)
0 + I. 1

O OCH3-?I-OCH3 . 7 CH2

OCH3 chain transfer I

(EH

OCH3—Si-OCH3

OCH3

O CH
3; 0 CH3

0

IN" c.

1 . . (11
«EH HOI VTMOS grafting (EH2

OCH3-Si-OCH3 9H2

OCH3-Si-OCH3
OCH3

 

 

Figure 6.5: Proposed reaction of VTMOS grafting onto PLA

138

 

 

 

0 CH3
h dro/ sis 0
"1* 11,8 1.19% )+ 381,81.
1%
9H2 _ [-2
H3CO-Si-OCH3 ”0 Sll OH
OCH3 OH
0 CH3 0 8., d .- 69.1.1129“ 9%. H. 0
811 M "——» “300:; .. -.
o 0
CH2 9H2 0” 0” \r/
CH2 CH2 crosslinks between chains
Ho—si—OH HO-SlI-OH
OH OH

 

Figure 6.6: Proposed reaction of moisture crosslinking in VTMOS-grafted PLA

Role of the Catalyst

The moisture crosslinking reactions take place spontaneously in the
presence of water, however, the process is very slow, making the aid of a
suitable catalyst of practical importance. Organotin catalysts are commonly
employed in silane grafting reactions to expedite the crosslinking process. In this
study, DBTDL was incorporated during the second extrusion pass. The exact
mechanism by which DBTDL accelerates the process is not fully understood, but
it is believed to be broken down to yield dibutyltin laurate hydroxide and Iauric
acid (Figure 6.7). The reaction products are what are believed to act as the
actual “catalyst”, whose overall role is to reduce the activation energy to
accelerate hydrolysis of the alkoxysilane groups to silanols followed by their

condensation to siloxane.33"58

139

 

 

 

 

 

0
M
0—0— (311st 0,119 o—c—c C11H23
H9\ sn/ He\ sn< 11Hza I
S"\ + H20 /Sn\ . —o
C4Hg l-IO—C—C11I-l23 C4l-l9 OH I
H OH
[lbutyltin dilaurate Dibutyltin laurate hydroxide Lauric acid

 

Figure 6.7 Expected mechanism of DBTDL reaction33

Effect of Silane Concentration on the Degree of Crosslinking

The degree of crosslinking was determined by measuring the gel content
of the grafted polymers. Figure 6.8 depicts the effect of silane concentration on
the samples prepared in the presence and absence of catalyst measured after
the maximum crosslinking period. The concentration of silane was varied from
0% to'3%, while the concentration of peroxide and catalyst were kept constant at
0.1% and 0.05%, respectively.

A greater degree of VTMOS grafting was expected to ensue more
crosslinking, and hence, a higher gel content. As both curves in Figure 6.8
indicate, increasing the VTMOS content indeed induced a rapid increase in the
degree of crosslinking at low VTMOS concentrations. At higher VTMOS
concentrations, the curves gradually leveled off to a constant value. This is prime
indication that the addition of VTMOS beyond a certain amount has very little
effect on the gel content. This is attributed to the possible homopolymerization of
the VTMOS resulting in long molecular chains of monomers, which would then

reduce the effective grafting yield of silane onto PLA.59 As expected, the

140

catalyzed samples performed better than the uncatalyzed samples. The
maximum gel contents were nearly 70% in the presence of catalyst and

approximately 58% in the absence of catalyst.

 

 

 

 

 

 

 

, ‘I ------- 1
3"
v T
E j
.9
C
§
8 —o—0% DBTDL

- - -0.5% DBTDL
0 2 4 6

 

 

VTMOS (wt%)

 

Figure 6.8: Effect of VTMOS concentration on the gel content of samples
prepared with 0.1% L101 after 28 days of crosslinking

Effect of Initiator Concentration on Gel Content

Figure 6.9 shows the inﬂuence of initiator concentration on the gel
contents in the catalyzed and uncatalyzed samples when VTMOS was held
constant at 3% and measured at the end of the 28day period. With increasing
initiator concentration similar trends were observed in the gelation behavior for
both sets of samples. The gel content was observed to initially increase sharply
with increasing initiator content, inferring a high rate of gel formation. Thereafter,
the rate decreased as the achievable gel content of 79% was reached at 0.2%
L101 and maintained upon further L101 addition. The initial increase in gel

content is due to primary free radicals generated by peroxide decomposition

141

either favoring chain combination or inducing more sites for grafting.
Consequently, more crosslinking would occur. At higher L101 concentrations no
further increase in gel content was seen in the range of the examined
concentration. This indicates that the maximum attainable amount of gel had
been achieved in both sets of samples in the cure period. Overall, these curves
show that the initiator had more inﬂuence on the gel formation than did the
VTMOS concentration, as evidenced by the difference in gel contents in the two

sets of experiments.

 

90
80
70
60
50
40
30
20
1 0

0

 

 

"0— 0% DBTDL
- I '0.05% DBTDL

Gel content(%)

 

 

 

 

0 0.2 0.4 0.6
L101 (wt%)

 

 

 

Figure 6.9: Effect of L101 concentration on the gel content of samples prepared
with 3% VTMOS after 28 days of crosslinking

Effect of Catalyst Concentration on Gel Content

The effect of catalyst concentration on the gel content was also studied to
identify the optimal catalyst concentration for achieving sufﬁcient crosslinking.
The sample composition investigated was that which was prepared with 0.1%

L101 and 3% VTMOS. Catalyst concentrations of 0.05%, 0.1%, and 0.2% were

142

used and the gel contents were measured after 28 days. The relationship
between gel content and catalyst concentration is depicted in Figure 6.10. Based
on the data, no signiﬁcant differences were observed in the samples, except for
the sample containing 0.2% catalyst in which the insoluble fraction dropped to
49%, an effect which could be attributed to increased degradation of the polymer,
producing fragmented polymer chains that are soluble in dichloromethane. These
results indicate that at low concentrations and after longer curing times the effect
of the catalyst on the gelation is miniscule, therefore, only a small concentration

of catalyst is necessary for achieving maximum attainable degree of crosslinking.

 

100

 

Gel Content (%)
8 8

 

 

O I I I I
0 0.05 0.1 0.15 0.2 0.25

Catalyst concentration (wt%)

 

 

 

 

Figure 6.10: Effect of catalyst concentration on the gel content of samples
prepared with 3% VTMOS and 0.1% L101 after 28 days of crosslinking

Effect of Crosslinking Time on Gel Content

The gel content was measured and plotted as a function of crosslinking
time. As Figure 6.11a illustrates, the gel contents of the samples prepared with

0.5% and 1% VTMOS reached a maximum of less than 10% and remained

143

constant throughout the entire crosslinking period. However, those of the 3% or
5% VTMOS samples increased with crosslinking time until about 21 days, when
the values began to level off to nearly 60%. When catalyst was added, as shown
in Figure 6.11b, the samples prepared with 3% and 5% VTMOS achieved a
higher gel content in a shorter time frame, which gradually increased to
approximately 70% at the end of the crosslinking period.

Figure 6.12a shows increasing gel content with increasing L101
concentrations and increasing time. The samples with catalyst in Figure 6.12b
illustrate less change in gel content over the crosslinking period. These data
imply that the relationship between crosslinking time and gel content is
dependent upon whether catalyst was added. In both Figures 6.11 and 6.12, it
can be seen that samples with added catalyst achieved the maximum gel content
in a shorter time frame than the samples without any catalyst, suggesting that the

catalyst addition can increase efﬁciency of the overall crosslinking process.

144

 

 

 

"'15:“ 0.5% VTMOS
"ﬁ- 1% VTMOS
'9' 3% VTMOS
+ 5% VTMOS

 

 

 

 

 

 

Gel content (%)

80

70 - (b)

60 _.

50 _ +05%vwos
40 ~ "ﬁr-1%V1MOS

30 _ .. -e-3%V1M0s
20 - / +5%vmos

0 10 20 30 40
Crosslinking time (days)

 

 

 

 

Figure 6.11: Effect of crosslinking time and VTMOS concentration on the gel
content of crosslinked samples (a) no DBTDL (b) 0.05% DBTDL

145

 

 

 

 

 

 

 

 

 

 

 

90
80 -1
70 1
+0% L101
60 1
+0.05% L101
50 ..
+0.1%L101
40 -
+02% L101
30 ..
-e-0.5% L101
20 1
10 -
0 -
390
g 80 . (b)
5
o 70 1
'3 +0%L101
(D 60 -
+0.05% L101
50 .
+0.1%L101
40 1
T +02% L101
/ -0-0.5% L101
20 1
10 - WV
0 -J:———"~.- ; . .
0 10 20 30 40
Crosslinking time (days)

 

 

 

Figure 6.12: Effect of crosslinking time and L101 concentration on the gel content
of crosslinked samples (a) no DBTDL (b) 0.05% DBTDL

146

Rate of Crosslinking

The rate of crosslinking of the samples prepared with and without catalyst
was calculated based on a method derived by Kelnar and Schatz (Equation

6.6):60

_ Ga _ Gt
r(t) — In [—00, _ Go] (6.6)

where G, is the gel content at time t, Ga is the maximum gel content, and GO is

the initial gel content. A plot of the logarithm of relative gel content as function of
crosslinking time is shown in Figure 6.13. The slope of the straight line was taken
as the “rate constant”, which is shown in Table 6.3. The rate constant for the
crosslinking reactions in the absence of DBTBL is slightly lower than that of the
sample in the presence of catalyst, which conﬁrms that DBTDL is capable of

accelerating the crosslinking reaction.

 

 

   

-
---
--
‘-
‘-
‘-
-‘
-
--

* y = -0.0876x + 0.1273

  

y= -0.1611x - 0.3154

Ln of Relative Gel

 

 

 

-3.0 1
-4.0 1 i
o 0% DBTDL
-5.0 1
I 0,,
-6.0 I0.05/ DBTDL 1 I I
0 5 10 15 20 25 30

Crosslinking time (hours)

 

 

 

Figure 6.13:Linear dependence of crosslinking time on the relative gel content of
samples prepared in the presence and absence of catalyst

147

Table 6.3: Crosslinking Rate Constants of Crosslinked Samples

 

Rate Constant (min'1)
No catalyst 0.001
0.05% DBDTL 0.003

 

 

 

 

 

 

Degree of Swelling

According to network theory, two crosslinks per polymer molecule are
sufﬁcient to crosslink a substantial portion of chains and increase the gel content
to its maximum. In other words, gel content is not necessarily an accurate
measure of the crosslink density, but rather the crosslinking ability. Polymer
networks with differing crosslink densities can result in the same degree of gel
formation. Therefore, in order to make quantitative inferences of the crosslink
density, another property must be considered—the equilibrium swelling ratio (Q).
Q is deﬁned as the volume ratio of a gel in its fully swollen state to that in its dry
state. In many practical cases, the reciprocal of the swelling ratio (1/Q) is—to a
good approximation—directly proportional to the crosslink density. The study of
the equilibrium swelling ratio of the crosslinked PLA indicated that the degree of
swelling was reduced with increasing L101, implying a tighter crosslinking
network, hence a higher crosslinking density. Figure 6.14 illustrates the
relationship between 1/Q and varying the initiator concentration after 28 days of
curing. Comparison of the swelling ratios with the gel contents shows that more
siloxane linkages were formed even after the attainable gel content was

achieved.

148

 

 

1/Q

 

 

 

 

0.02 1 —o-1/Q—0% DBDTL
0-01 ‘ —a—1/o—0.05% DBDTL
0 I l r I l
0 0.1 0.2 0.3 0.4 0.5 0.6
L101 (wt%)

 

 

 

Figure 6.14: Reciprocal swelling ratio as a function of initiator concentration
determined after 28 days of crosslinking

Proof of Grafting by XPS

XPS was used to verify the presence of silicon in the crosslinked samples.
Figure 6.15 displays the spectrum of PLA. Figures 6.16 through 6.18 present
XPS spectra of the crosslinked PLA prepared with 0.1% L101 and 0.05% DBTDL
with varying VTMOS concentrations. Figures 6.16 through 6.18 all reveal that the
crosslinked PLA has additional peaks at 103 eV and 154 eV when compared to
the unmodiﬁed PLA. The peak at 103 eV corresponds to Si2p and that at 154 eV
is representative of Si2s. The XPS data conﬁrm the presence of VTMOS

grafting/crosslinking in the tested samples.

149

 

 

 

 

 

10.E+O4
(I)
5 :2
. _ O
08.15104 j , '
_. x I 5
-' O
06-E+04 1 i", 5 I I
.. u. r
I.\. ".31“ .'| 5 t
0" " ”up“ . S ‘
O4.E+O4 " .0". L»‘Q"'v“' S
'J
"1...... .. ...rou'é
02-E+04 1 ’I
"\cgo.ocoa-O--o..‘
OO.E+OO
1100 1GB g1) 8(D 700 600 3» 4CD III) Z!) 10) 0
BindngEneIgi/(GV)

 

 

 

Figure 6.15: XPS spectrum of extruded PLA

 

 

 

 

 

 

 

12E+04
3
O
10E+04 1 ,
:1
08E+04 - _J x f
5 C: = .
06E+04 1 L," .55"; . 5
"1"". 5 5.5.
04E+04 ‘ W23. 5
". ,:1 Q g
"wM’MEE % (7)
02E+04 1 $5
..NQ-u—nuLhL-A'N"
00E+00 .
11001m0w08007006005104m300301000
BirmEnerrwleVI

 

Figure 6.16: XPS spectrum of crosslinked PLA prepared with 0.5% VTMOS

150

 

 

 

 

 

 

 

 

 

12.E+04
8
10.E+04 1 0
1i
08.E+04 ~ g 0 g
Q
06.E+O4 1
+ u D.
04.5 04 I 33 I};
02.E+04 5 ELL...
00.E+00
110010009008007006005004003002001000
BindngEnergylevl

 

Figure 6.17: XPS spectrum of crosslinked PLA prepared with 1% VTMOS

 

 

 

 

 

 

 

10504
08E+04 - E
O
.J
_J
06E+04 1 .1 x
2’ ° .2
O O
04E+04 -
02E+04 ; T
00E+00,
11001000 900 800 700 600 500400 300200100 0
Emma EnergerV)

 

 

 

Figure 6.18: XPS spectrum of crosslinked PLA prepared with 3% VTMOS

151

Effect of Crosslinking on Melt Flow Behavior

As expected, crosslinking had an appreciable effect on the melt behavior
of PLA, as shown by the melt ﬂow index (MFI) data in Table 6.4. The addition of
VTMOS decreased polymer ﬂow, eventually to levels where the MFI could not be
determined due to the highly crosslinked network. Likewise, increasing the
peroxide content also resulted in the greater decrement of MFI, that is, until 3
L101 concentration of 0.5% was reached, which increased the MFI to a
measurable value. It was apparent that there was competition between grafting
and chain scission and the increase in ﬂow is believed to be a result of
degradation by chain scission. Typically, the addition of excess VTMOS relative
to the peroxide content will ensure that grafting prevails over chain scission.”
However, the melt ﬂow behavior attests that at this L101 concentration, chain
scission prevailed in the linear portion of the polymer. In the highly crosslinked
samples, including all of those prepared with catalyst, crosslinking reactions were

dominant, preventing ﬂow of the polymer even at elevated temperatures.

152

Table 6.4: Melt Flow Index Values for Crosslinked PLA

 

Sample L101 VTMOS DBTDL MF I

 

 

# (%) (%) (%) (9110mm)
1'“ 0 0 - 13.0 1 1.7
2 0 O -- 15.8 1 1.2
3 0.1 0 -- 14.2 1 1.5
4 0.1 0.5 -- 12.4 1 2.2
5... 0.1 1 -- 10.01 1.1
6 ... 0.1 3 -- -
7 0.1 5 -- -
8 ... 0.05 3 -- 1.7 1 0.7
9 0.2 3 - -
10... 0.5 3 -- 1.8 1 0.1
11 ... 0.05 3 0.05 —
12 ... 0.1 3 0.05 --
13 .. 0.2 3 0.05 --
- 14 0.5 3 0.05 --
nneat PLA
extruded PLA

... would not melt at 190°C or 230°C

Thermal Properties

Tables 6.5 and 6.6 show the DSC data of the crosslinked samples as a
function of L101 and VTMOS content, respectively. One observation that must be
noted is, despite identical heating conditions, the. samples that were loaded with
3% VMTOS all displayed only one melting endothenn, while those with lower
VTMOS loadings showed two. This indicates the existence of a level of VTMOS,
above which signiﬁcant crosslinking and changes in morphology occur. In this
case, that level is 3%. The presence of multiple endotherms is an effect that is
often seen in PLA-based polymers and is usually a result of the melting of
imperfect, lower melting crystals then their recrystallization into higher-melting,
more perfect crystals. The disappearance of the higher temperature endothenn

in the 3% VTMOS samples may be attributed to the presence of only one

153

crystalline “type” due to an increased amorphous phase by the incorporation of
crosslinks. This effect may also account for the drop in crystallinity that was seen
in these samples. Crosslinks work to impede mobility, preventing chains from
ordering into a crystalline structure. As a result, the samples containing more
crosslinks have fewer linear chains that are able to crystallize. The samples with
a constant peroxide concentration and varying VTMOS concentrations (up to 3%
VTMOS) are less crosslinked and therefore have more linear chains excluded
from the polymer network. These chains are more likely to recrystallize into
different types of crystals, thus explaining the multiple endotherms.

The glass transition, melting, and crystallization temperatures remained
fairly constant throughout all sample compositions. The heats of fusion and
percent crystallinities of the samples are also shown. The samples with constant
VTMOS concentration showed lower crystallinities than that of the neat polymer,
while those initiated with a constant amount of L101 experienced no signiﬁcant
change, except for the sample prepared with 3% VTMOS. The lower degree of
crystallization suggests that the molecular chains were crosslinked in the
amorphous state.32 Those crosslinks act as defect centers to suppress the
molecular motion and impede chain folding required for crystallization, hence the

decrease in crystallinity and melting enthalpies of the samples.

154

._o>o_ mocmoccoo $3 05 um moocmeotﬁ E8536 E3059 cE:_oo 38mm 05 £55, Beam. Eoemtﬁ

 

 

81.9.1.9 8185.8 -- 833.: 635:: -- 53.3w: «1.0.3.3 to m
5.33.0.8. 8.0.3.2 8233.? «more 538.33 50.53.83 11.31.03 53.9.15 to 5
833.2 1315.8 1331 No.33 116332 638.va 58.05.53 1333 to no
6531.? 133.: 538.9 63.8.». 538.8 @3533 5035.13 «138.3 to 0
53.5.8 aware... mode: and“: 133.02 «~58? «928.13 533.8 0 o
xe\e A093 waa 3:3 Away 68v 68v AWOV 38v 3%
1a Ia Fara e we; .5» e 83 wo_E.>

 

22511880 8.25 BE? 55, <._n_ 89.5.8820 .5 28 own 88 531p

._o>m_ mocwoccoo $3 05 E 805551628556 E6859 5:28 9:3 or: £53, Bose. ESQ—to

 

 

nmdﬁvdw nmdﬂhd I omhdﬂvd modﬂwdww I nmNdH v.03. whdﬂﬁwm m6 m
315.: oedema -- 838.2 135:: -- when}: 5338 No m
n0.0H©.N—. nwdﬁmdv l omdﬁﬁ. 2. mm. FHN. —. S. I no. FHONVr $06..“me _..o m
mhdﬂmﬁw no. _.Ho.m l nhdﬂ—mvr nQ.NHm.NN_. l mvdﬂwdvv mmdﬂvﬁm mod m
135.? 555.3 SEE and“: 533.12 «.589 «1.31.1: 10.5.8 o o
S. mum «Emma amuse 8.1.; Me...” 6.3 2.; 6.5 may 3%;

 

825.2880 55:2. 855$ 5? Se 815.890 .5 man own ”no 633

155

The DSC data of the samples prepared with catalyst are shown in Table 6.7. It

was apparent that the addition of catalyst had no major effect on the T9 until the

catalyst amount was raised to 0.2%, at which point the Tg decreased to nearly

52°C. At a catalyst concentration of 0.05%, the crystallinity was signiﬁcantly lower
than that of the other samples. Upon further catalyst addition, the degree of

crystallinity progressively increased to nearly 24% at a catalyst addition of 0.2%.

This, along with the decline in T9, is attributed to the degradation caused by

excess catalyst, with the resulting fragmented chains acting as plasticizers in the

polymer bulk.

156

157

._m>o_ mocmnzcoo $3 05 .m woocmcmtﬁ E8536 E0359 c828 9:3 on. 55:5 90:2 E9030

 

 

cinemm $33.: umﬁoem 90:33 £33m: .3395 No to m
comm: “0:3 swim? sadism? 59%an “3&5 to to m
1.33 neonate among neoaoaﬁ 923.2: $43.8 mod to m
«echoes @938 muons: “mete: “33er meoamsm o to m
85 6v 8v 3 A0: A3
950 O o o o o o o o
\c :3 ma 2% 6% F as mp 895 53 8.2;

 

289.6880 $298 9.3:; 5? Sn. Bxsameo do 28 own Ho 2an

Thermal Stability

One of the primary objectives of crosslinking polymers, PLA in particular,
is to increase the upper service temperature limit by restricting chain mobility and
preventing thermal degradation when heated at or above the crystalline melting
temperature. An investigation of the thermal stability of the crosslinked products
was performed using TGA. The thermogravimetric results of neat PLA and
VTMOS crosslinked PLA, with and without catalyst, are shown in Figures 6.19
and 6.20, respectively. As can be clearly seen from the ﬁgures, both catalyzed
and uncatalyzed samples display similar trends. All of the crosslinked samples
showed improved thermal stability over the neat polymer, with the exception of
the sample initiated with 0.5% L101 due to its molecular weight reduction by
chain scission. Figure 6.20 shows that the catalyzed samples were only slightly
more thermally stable than the uncatalyzed ones. This increase in decomposition
temperature could be due to a higher degree of crosslinking in this system,

previously shown by increased gel content.

158

 

 

 

 

 

 

 

......... PLA
100 -. ---D.05% L101
$563. - ~0.1%L101
_ '--._x 6A — -- 0.2% L101
33 80 33‘3". ——U.5% L101
E 60 -
O)
‘65
g 40 _
20 -
0
320 330 340 350 360

Temperature (°C)

 

 

Figure 6.19: TGA traces of crosslinked PLA with varying initiator concentrations

 

 

 

 
 
   

120
......... PLA
100 —r-..«..-;,-,~.,,=;;_-an ----- 0.05% L101
--._~~\_ -- —u.1% L101
80 ~ “'9‘- — - - 0.2% L101
——u.5% L101

Weight (%)
O)
O

 

 

 

320 325 330 335 340 345 350 355 360
Temperature (°C)

 

 

Figure 6.20: TGA traces of crosslinked PLA with varying initiator concentrations
prepared with 0.05% catalyst

159

Improvements in the thermal stability of PLA resulting from crosslinking
prompted investigation of the inﬂuence of catalyst concentration on the thermal
decomposition. While holding constant the initiator and VTMOS concentrations,
the catalyst content was increased from 0.05% to 0.2% and the corresponding
thennograms are presented in Figure 6.21. An increase in catalyst concentration
did not increase the thermal stability as was expected, rather it decreased it
signiﬁcantly. This is most likely a result of the depolymerization of PLA with
excess catalyst, providing further evidence that only a small amount of catalyst is

necessary for achieving sufﬁcient crosslinking.

 

 

 

 

 

 

 

120
- - - PLA
— - - no catalyst
100 - . :m‘x, — - 0.05% catalyst
.......... \ 0.1% catalyst
80 ~
3
E 60 -
.9
(D
g 40 .
20 6
0 l
280 300 320 340 360
Temperature (°C)

 

 

 

Figure 6.21: TGA traces of crosslinked PLA with varying catalyst concentrations
at a constant 3% VTMOS and 0.1% L101 loadings.

The TGA data show evidence of correlation between the thermal stability

and crosslink density of the crosslinked samples indicating that the crosslinking

160

reaction of VTMOS-grafted PLA could possibly raise its upper limit temperature
of application. Similar results have been reported in the literature with other

polymer systems.32

Mechanical Properties

The mechanical behavior of the crosslinked polymers was investigated
and is illustrated in Figure 6.22. PLA itself is a fairly brittle polymer with high
strength and modulus. Crosslinking was expected to have an effect on the
mechanical behavior of PLA because the strength and modulus of semi-
crystalline polymers are dominated by the hard crystalline regions which are held
together by strong forces. With the addition of crosslinks, the polymer is unable
to crystallize as easily, weakening the forces holding together the crystalline
structure, usually reducing its strength and modulus. In this instance, the tensile
strength and modulus remained relatively unchanged as a result of crosslinking.
Since PLA remains partially crystalline even after crosslinking, the crystallites in
the original polymer are responsible for the lack of any signiﬁcant changes in the
tensile properties.

In most cases, the elongation at break is also affected by crosslinks
between chains because they tend to impose a greater restriction on the
elongational properties, due to shorter segments in the polymer. This prevents
chain entanglements and slippage, thereby resulting in a reduction in elongation
at break. Nevertheless, in this work a reduced elongation was not seen.
Interestingly, in the 3% VTMOS sample an increase in elongation was seen (from

4.3% to approximately 8%). This particular sample also had the highest gel

161

content and the lowest measured crystallinity. It should be noted that in the highly
crosslinked samples the properties are mainly governed by the chemically
introduced crosslinks and the VTMOS grafted chain ends that are not excluded

from the crosslinked network.

 

lTensile Modulus I Tensile Strength I Elongation at Break

  
  

 

 

 

 

1)
CA
1; 4000 7 - 7o 25:
“2' 6°
3 3000 ~ - 50 593
,3, 2000 . :3 g;
- O
:5 1000 . 7735
c ' 10 gm
'9 0 . c 0 '-
s s s
9U“ WM“- 9U"\ 9U“

 

 

 

Figure 6.22: Mechanical properties of crosslinked PLA prepared with varying
VTMOS concentrations and 0.1% L101 with 0.05% catalyst.

CONCLUSIONS

It can be concluded that VTMOS can be effectively grafted onto
commercial grade PLA via a free radical initiation and the resultant polymer can
successfully undergo moisture crosslinking to form siloxane linkages. Evidence
of the presence of silicon moieties in the polymer was provided by XPS analysis.
The effects of monomer, initiator, and catalyst concentrations on the reaction
products indicated the composition corresponding to the optimum properties and

showed that presence of only a small amount of catalyst showed to be a major

162

contributor to the crosslinking formation in the time frame investigated.
Furthermore, thermogravimetric data showed an increase in thermal stability as a
result of crosslinking. An investigation of the mechanical properties revealed that
this crosslinking technique enabled PLA to retain its mechanical strength without

resulting in polymer embrittlement.

163

REFERENCES

1. Takamura, M., Nakamura, T., Takahashi, T., Koyama, K., Poly. Degrad.
Stab, 2008, 93, 1909.

2. Di, Y., lannance, S., Di Maio, E., Nicolais, L., Macomol. Mater. Eng., 2005,
290, 1083.

3. Shinno, K., Miyamoto, M., Rimura, Y., J. Appl. Polym. Sci., 2003, 90,
1802.

4. Zhong, W., Ge, J., Gu, Z., Li, W., Chen, X., Chen, Y., Yang, Y., J. Appl.
Polym. Sci., 1999, 74,2546. .

5. Hiltunen, J., Seppala, V., Harkonen, M., J. Appl. Polym. Sci., 1997, 63,
1091.

6. Klyma, J., Tuominen, J., Helminen, A., Seppala, J., Polymer, 2001, 42,
3333.

7. Garlotta, D., J. Polm Envir, 9, 2, 2001, 63.

8. Khonakdar, H.A., Morshedian, J., Wagenknecht, U., Jafari, S.H., Polymer,
44, 2003, 4301.

9. Nijenhuis, A.J., Grijpma, D.W., Pennings, A.J., Polymer 1996, 37, 2783.

10.Hartmann, M.H., In: D.L. Kaplan, Editor, Biopolymers from renewable
resources; Springer-Verlag: Berlin, 1998.

11.Gruber P.R., Hartmann M.H., Kolstad J. J., Witzke D. R., Brosch A. L.,
Cargill, Method of Crosslinking PLA, USA, PCT94/08 508, 1996.

12.Ryan C. M., Hartmann M. H., Nangeroni J. F., In: Polymers, laminations,
and coatings conference, TAPPI Proceedings.1997, 139.

13. Babanalbandi, A., Hill, D., Whittaler, A., Polym. Degrad. Stab., 1997, 58,
203.

14.Burdick, D. Frankel, S. Demall, Anseth, K.S., Biomaterials, 2003, 24
1613.

15.Mimoto, H., Kaneda, A., Quynh, T.M., Nagasawa N., Yoshii, F., Polymer,
2005, 46, 4695.

164

16.Sodergard A., Nasman, J.H., Ponm. Degrad. Stab., 1994, 46, 25.
17.Sodergard A., Nasman, J.H., Ind. Eng. Chem. Res., 1996, 35, 732.

18.Sodergard, A., Niemi, M., Selin, J.F., Nasman, J.H., Ind. Eng. Chem. Res.,
1995, 34, 1203.

19.Dorgan, J.R., Lehennier, H., Mang, M., J. Polym. Envir., 2000, 8, 1.

20.Raquez, Jean-Marie, Narayan, R., Dubois, P., Macrom. Mater. Eng.,
2008, 293, 447.

21. Carlson, D., Dubois, P., Nie, L., Narayan, R., Polym. Eng. Sci., 1998, 38,
2, 311.

22.Hartmann, M.H., High Molecular Weight Polylactic Acid Polymers;
Springer Verlag: Berlin, 367.

23.Sodergard, A., Selin, J.-F., Niemi, M., Johansson, C.-J., Meinander, K.,
US. Patent 6559244, 2003.

24.Carlson, 0., MS. Thesis, Michigan State University, East Lansing, MI,
1995.

25. Han, C., Ran, X., Su, X., Zhang, K., Liu, N, Dong, L., Polym. Int., 2007,
56, 5, 593.

26.Bradley, R., Radiation Technology Handbook; Informa Health Care, 1984.

27.Nugroho, P., Mitomo, H., Yoshii, F., Kume, T., Polym. Degrad. Stab.,
2001, 72, 337.

28.Banana|bandi, A., Hill, D., O’Donnell, J., Pomery, P., Whittaker, A., Polym.
Degrad. Stab., 1995, 50, 297.

29.Loo, S.C.J., Ooi, C.P., Boey, Y.C.F., Polym. Deg. Stab., 2004 83, 259.

30. Nagasawa, N., Kaneda, A., Kanazawa, S., Yagi, T., Mitomo, H., Yoshii, F.,
Tamada, M., Nuc. Inst. Meth. Phys. Res. Part B., 2005, 23, 6, 611.

31.Jin, F., Hyon, S.H., Iwata, H., Tsutsumi, S., Macromol. Rap. Commun.,
2002, 23, 909.

32.Quynh, T., Mitomo, H., Nagasawa, N., Wada, Y., Yoshii, F., Tamada, M.,
Euro. Polym. J., 2007, 43, 1779.

33.AI-Malaika, 8., Ed. Reactive Modiﬁers for Polymers; Chapman & Hall:
Cambridge, 1997.

165

34.Scott, H.G., US. Patent 3646155, Crosslinking of a Polyoleﬁn with a
Silane, 1972.

35.Wu, T-S., Alchemix Pte Ltd., Plas. Add. Comp. 2007.

36.8warbrick, P., et al., Manufacture of Extruded Products, US. Patent
4117195, 1975.

37.Sen, A.K., Mukherjee, B., Bhattacharyya, A. S., De. P. P., Bhowmick,
A.K., J. Appl. Polym. Sci., 1992, 44, 1153.

38.Bullen, D. J., capaccio, G., Frye, C. J., Brock, T., Bri. Polym. J., 1989, 21,
117.

39.Narkis, M., Tzur, A., Vaxman, A., Polym. En. Sci., 1985, 25, 857.
40.Shieh, Y.-T., Tsai, T. —H., J. Appl. Polym. Sci. 1999, 69, 255.
41.Shieh, Y.-T., Liu, C.-M., J. Appl. Polym. Sci., 1999, 74, 3404.

42.Sen, A.K., Mukherjee, B., Bhattacharyya, A. S., De. P. P., Bhowmick,
A.K., J. Thermal Anal. 1993, 39. 887.

43.Turcsanyi, B., Fekete, E., Pukanszky, B., Tabor, I., J. Thermal Anal.,
1990, 36, 1775.

44. Kao, Y.H., Phillips, P., J., Polymer, 1986, 27, 1669.
45.Shah, G. B., Fuzail, M., Anwar, J., J. Appl. Poly. Sci., 92, 2004, 3796.
46. Kelnar, l., Schatz, M., J. Appl. Polym. Sci., 1993, 48, 657.

47.Hidalgo, M., Beltra' n, M. L., Reinecke, H., Mijangos, C., J. Appl. Polym.
Sci., 1998, 70, 865.

48.Nachtigall, S. M. B., Stedile, F. C., Felix, A. H. 0., Mauler, R. S., J. Appl.
Polym. Sci., 1999, 72, 1313.

49.Yang, 8., Song, G., Zhao, Y., Yang, C., She, X., Polym. Eng. Sci., 47,
2007, 1004.

50.Huang, H., Lu, H. H., Liu, N.C., J. Appl. Polym. Sci., 78, 2000, 1233.
51.Liu, N. C., Yao, G.P., Huang, H., Polymer, 41., 2000, 4537.

52.Zhang, G., Wang, G., Zhang, J., Wei, P., Jiang, P., J. Appl. Polym. Sci.,
2006, 102, 5057.

166

53.Giiven, 0., Proceedings of the NATO Advanced Study Institute on
Advanced Methods of Determination of Crosslinking and Scission in
Polymers and Their Effects on Mechanical Properties Held in Kemer-
Antalya, Turkey, September 4—16, 1988.

54.Stevens, M., Polymer Chemistry: An Introduction; Oxford University Press
US, 1999.

55.Swelling measurements of crosslinked polymers; Cambridge Polymer
Group Inc: Boston, MA, CPGAN 005, www.campoly.com

56. Balakrishnan, 8., Ph.D. Dissertation, Michigan State University, 2004.

57. Sperling, L.H., Introduction to Physical Polymer Science, 3rd ed; Wiley:
Hoboken, NJ, 2001.

58.Shah, G.B., Winter, R.W., J. Appl. Polym. Sci., 1996, 61, 1649.
59.Yussuf, A. A., Kosior, E., Alban, L., Malaysian Polym. J., 2007, 2, 2, 58.
60.Kelnar, l., Schatz, M., J. Appl. Polym. Sci., 1993, 48, 4, 668.

167

Chapter 7

 

REACTIVE COMPATIBILIZATION OF POLYLACTIC ACID AND

POLYBUTYLENE ADIPATE CO-TEREPHTHLATE

 

INTRODUCTION

In recent years, increased interest in effectively managing global carbon
resources and reducing our carbon footprints has sparked the development of
compostable and biobased polymers for single-use and disposable products,
speciﬁcally for blown ﬁlm applications. Of the biobased polymers currently on the
market, polylactic acid (PLA) is, for many reasons, a very attractive option.
Notwithstanding the high strength and stiffness that PLA offers, its high T9 and
susceptibility to physical aging render it hard and brittle. This, along with its low
melt strength at elevated temperatures cause PLA to fall short of the necessary

properties required for blown ﬁlm manufacturing.

168

in:

DE

SI

9“

pr

Pl

Combining PLA with ﬂexible polymers appears to be the natural solution to
combat some of these issues, while extending its range of achievable properties.
Polymer blending is a well-established strategy to improve properties of
individual polymers and is becoming increasingly important in polymer
production. Just as novel polymers are being developed from the polymerization
of individual monomer units, the blending of two or more existing polymers can
lead to new materials with unique properties. In doing so, properties of the
individual components can be retained and combined or modiﬁed to collectively
meet the various demands of the resulting polymer.1

Blending polymers with PLA represents alternate routes to develop
partially biobased materials with tailored properties, which are intermediate to or
synergistically favorable to over the component homopolymers. Various
properties of these polymer blends are particularly interesting as they can be
related to the interactions between component phases, phase morphology, and
processing parameters.

In the following sections, blends that have been produced from PLA and
other polymers will be discussed. In addition, theory pertinent to polymer

blending will be presented.

PLA-Based Blends

PLA has been blended with numerous different types of polymers
(synthetic, organic, miscible, and immiscible) mainly for the reason of improving
the ductility of PLA. From the literature, it has been found that only a limited

number of polymers are able to form miscible blends with PLA. For example, low

169

molecular weight polyglycols are miscible with PLA at low concentrations acting
as polymeric plasticizers. However, higher concentrations of polyethylene glycol
(PEG), according to Sheth et al.2, contribute to a decrease in modulus and
elongation at break along with lower degradation temperatures. In another work,
Nijenhuis et al.3 conﬁrm these results and attributes the increased degradation to
the hydrophilic nature of PEG. Polyvinyl acetate (PVAc) is another polymer that
shows miscibility with PLA by the detection of a single Tg. Gajria et al.4 show that
the addition of low concentrations of PVAc (5-30%) can increase the tensile
properties to values higher than that of the pure PLA. Furthermore, studies of
enzymatic degradation showed that PVAc decreases the rate of degradation of
PLA. Because of this, Park and Im5 partially hydrolyzed PVAc into poly (vinyl
acetate-co-vinyl alcohol) poly (VAc-co-VA) and studied it in blends with PLA,
ﬁnding that the two polymers were miscible, but only up to 10% poly (VAC-CO-
VA), after which two distinct Tgs were observed.

Another family of polymers that is reportedly miscible with PLA is
methacrylates. Eguiburu et al.6 prepared blends of PDLA and PLLA, each with
poly (methyl methacrylate) (PMMA) and poly (methyl acrylate) (PMA). The
blends containing amorphous PDLA exhibited miscibility over the entire
composition interval presenting only one Tg. On the other hand, the blend
systems with semi-crystalline PLLA only presented some degree of miscibility in

the melt, undergoing phase separation upon cooling.

170

Biodegradable PLA-Based Blends
Since PLA is a biodegradable polymer, blending other biodegradable

polymers with PLA is a highly desirable means to alleviate some of the
environmental concern regarding single-use plastics in disposable product
applications. The development of a completely biodegradable blend is important
to combat some of the major problems associated with the ultimate disposal of
polymeric materials. Therefore, polymers such as thermoplastic starch7,
collagene, and biodegradable aliphatic polyesters and copolyesters have each
been blended with PLA to achieve a polymer blend that is entirely biodegradable.
Lee and Lee9 developed blends with PLA and polybutylene succinate adipate
(PBSA) and reported properties characteristic of an incompatible blend, yielding
poor tensile strengths and moduli. The strains at break and impact strengths,
however, were favorably affected by the addition of PBSA, even at low
concentrations.

Blends of PLA with polyhydroxybutyrate-co-hydroxyvalerate (PHBV) were

L10 and lannace et al..11 Although, it was possible for

investigated by Ferreira eta
interaction to have occurred between the two phases in both studies, the blends
were immiscible as evidenced by discrete phase separation and the presence of
multiple glass transitions. In addition, a direct correlation between crystallization
temperature and degradation time was established, showing that the crystallinity
of the blends decreased in relation to degradation time, while the crystallinity of

neat PLA was seen to increase. The reason for this is that the presence of PHBV

in the blend disrupted the crystallization of PLA.

171

Simple blends of PLA and polybutylene adipate co-terephthalate (PBAT),
an aromatic, aliphatic polyester based on terephthalic acid, adipic acid, 1,4-
butanediol, and modular units, have also been studied. Jiang et al. described
enhanced elongation and toughness with increased PBAT addition, however, at
the expense of the tensile strength and modulus.“13 Also with the addition of
PBAT improvements in melt elasticity and viscosity were seen. A patent issued to
Cortec, in which blown ﬁlms were produced, reports tensile strengths up to 6940
psi and elongation values of up to 360% for blends containing 5% PLA and 40%
PLA, respectively.14

One of the most notable polymer blend systems, however, is that
comprising of PLA and polycaprolactone (PCL). This polymer pair has received
much interest recently, especially in drug delivery systems where the need for a
strong, tough biocompatible polymer exists. Miscibility, morphological, and
mechanical studies have been performed on these blends and have shown
immiscible phases at elevated temperatures”, along with decreased tensile
behavior with PCL addition. Results also showed that the mechanical properties
and morphology can be altered by varying the composition, which can be
explained by the percolation theory. ‘6

However, given that most polymer pairs are intrinsically incompatible and
form immiscible blends, their simple melt blending usually results in poor
properties. The development of a blend with sufﬁcient properties depends on the
ability to control its morphology and interfacial adhesion. By reducing the

components’ interfacial tensions, one is able to produce an emulsifying effect,

172

leading to good dispersion of one phase in the other. Likewise, the generation of
strong bonds at the interface can, in principle, improve the stress transfer
between component phases and result in improved properties. In this respect,
extensive studies of the compatibilization of PLA and biodegradable polymers
using coupling agents were carried out by several authors. Some employed the

"'19, while others

use of reactive additives, such as peroxides and catalysts
prepared di- or tri-block copolymers to be utilized as compatibilizers at the
interface. Tsuji et al.20 found that a di-block copolymer of lactic acid and
caprolactone was miscible with both PLLA and PCL and improved the tensile
strength, modulus, and elongation at break. Conversely, in a similar set of
experiments on the mechanical behavior, Hiljanen-Vainio et al.21 reported that
while the copolymer improved processibility and the strain at break, it decreased
tensile strength and modulus. Nevertheless, the compatibilizer addition
accelerated the degradation of PLLA. Dell’Erba et al.22 studied the thermal and
morphological effects of a triblock copolymer on the blend, reporting a reduced
domain size, signifying an emulsifying effect on the system.

Although, the blending of PLA and PBAT has been reported12‘14, no
systematic investigation of the properties as they pertain to phase interaction has
been performed nor have any studies focused on the compatibility of the system.
While PBAT and PLA are both biodegradable polyesters, PBAT is not derived
from biobased feedstock. Therefore, blending PLA and PBAT not only can

potentially result in a product that performs well, it can also be a practical

strategy for imparting biobased content into PBAT. In this project, PLA was

173

reactively blended with PBAT and blown ﬁlm was produced. The high ductility
and toughness of PBAT permit the blend to be easily processed using standard
ﬁlm blowing apparatuses. The combination of the high strength and stiffness of
PLA with the ﬂexibility of PBAT can lead to a polymer capable of being blown into
ﬁlm resulting in synergistic properties, providing that they have sufﬁcient
interaction at their interface.

A method to obtain sufﬁcient interaction in a phase-separated polymer
blend is to add an interfacial modiﬁer to the blend that can enhance the adhesion
between the two phases and improve the properties of the resultant mixture.23
One type of interfacial modiﬁer that has shown signiﬁcant promise is a copolymer
that consists of the two monomers that comprise the individual homopolymers
that are being blended. Additionally, it is well known that two polyesters can
undergo a transesteriﬁcation reaction in which the conversion of one or two
esters into one or two other esters proceeds by reacting in the presence of heat
and an alcohol, a carboxylic acid, or a catalyst.24 The reaction yields a copolymer,
comprising of components of the individual polyesters.

In this work, reactive extrusion was employed to improve the interaction
between PLA and PBAT via a transesteriﬁcation reaction. The focus of this
project was on the thermal, mechanical, and morphological properties of
PBAT/PLA blown ﬁlms. The goal was to understand the effects of the coupling
reaction on the structure of the blend and how it relates to the overall blend

pedbnnance.

174

THEORETICAL BACKGROUND

Basic Thermodynamics of Polymer Blending

As was previously mentioned, not all polymer blends are miscible. In fact,
the most polymer blends are completely immiscible or only show partial
miscibility. These blends are characterized by a coarse, meta-stable morphology
with poor adhesion between phases and poor performance. The best way to
understand the basis of polymer blending phenomena is to consider
thermodynamics, which ultimately deﬁnes miscibility given in terms of Gibbs free
1:25

energy of mixing, shown in Equation 7.

AGmix :AHmix TTASmix (7-1)
where Hmix is the enthalpy of mixing, Smix is the entropy of mixing, and T is

temperature. It is this equation that governs the existence of a miscible or

immiscible polymer mixture. To have a system with complete miscibility AGml-x

must be negative, implying that for exothermic reactions spontaneous mixing will
occur. For endothermic reactions only high temperatures will provide the driving
force for miscibility.

The requirement that the Gibb’s free energy be negative is not the only
requisite in achieving miscibility because many immiscible blends can also have

a value of AGmix less than zero. For further understanding, other necessary

criteria for complete thermodynamic miscibility will be discussed herein.
Considering the Flory-Huggins theory of polymer solutions as it can be applied to

polymer mixtures, the Gibbs free energy can be expressed by Equation 7.2:26

175

 

AGml-x {Ry M914] 1n¢A {£51} In ¢B +2.13 MB] (7.2)
r xA 3‘3

where V, is a reference volume, taken as the molar volume of the smallest
polymer repeat unit. ¢A and ¢B are the respective volume fractions of polymers
A and B. x A and x3 are the respective degrees of polymerization of polymers A

and B. 1A3 is related to the enthalpy of interaction of the polymer repeat units,

each having a molar volume of V,.

By taking the second and third derivatives of AG with respect to the
respective volume fractions and setting them both to zero (Equation 7.3) one can
solve for the conditions of 1,43 (Equation 7.4) and the volume fractions (Equations
7.5 and 7.6) for achieving complete thermodynamic miscibility and
thermodynamic stability. Meeting these conditions will indicate a homogeneous

phase in a binary mixture and ensure against phase segregation. 26

aZAGmix _ a3AGm,-x

 

 

 

 

_ =0 (7.3)
646,12 6315,13
2
1 1 1
(ZAB) =- + (7.4)
C’ 2 mm xBuz
1/2
XB
WA)” _ 1/2 1/2 (7'5)
xA +xB
( ) xBl/Z
m = (7.6)
cr xA1/2+x 1/2

176

Miscibility requires polymer mixtures to exhibit a single phase over a range
of temperatures, pressures, and compositions and can be best illustrated by a
phase diagram. A general view of liquid-liquid phase equilibrium is shown in
Figure 7.1 (a) presenting the upper critical solution temperature (UCST) and the
lower critical solution temperature (LCST). The UCST is convex upward and
associated with the occurrence of an initially homogenous polymer mixture that
undergoes phase separation upon lowering of temperature. When a mixture
phase separates upon raising the temperature, the critical temperature is referred
to as the LCST and is concave upward. Most polymer systems exhibit only an
LCST because the polymer mixture generally passes through the glass transition
as the temperature is lowered. If a blend is held at a temperature below the
LCST, the blend will presumably be frozen in the single phase. At higher
temperatures the blend components have increased molecular mobility and tend
to dissociate, breaking into separate phases.27

Although both the UCST and the LCST can be used to describe the
generalities of thermal phase changes in polymer mixtures, in most polymer
mixtures of utmost relevance is the solid-liquid equilibrium behavior as shown in
Figure 7.1 (b). In such a system, polymers display two solid phases below the T9
line. At temperatures above the glass transition amorphous and glassy phases

exist. Above the T9 line only a liquid mixed phase is present.

177

 

Liquid — Liquid Phase ' SOIid - Liquid Phase
Diagram Diagram

3868/ melt

 

 

 
  
 
   
  

 

 

 

 

 

 

0
5
‘5
‘6 1 phase melt + crystal
0
g UCST
'— /\ T9

2 phases lass + c stal

(Cl 9 W (D)
4> 1’

 

 

 

Figure 7.1: Phase diagrams showing transition behavior in polymer blends 26

Miscibility Determination

Aside from the thermodynamic requisite of displaying negative Gibbs free
energy, physical detection can also provide supplementary evidence of a
miscible blend. Preparation of a thin ﬁlm from miscible amorphous
homopolymers should yield no macroscopic heterogeneity and, therefore, result
in an apparent transparent material. However, it is important to point out that
transparent ﬁlms can also be formed by amorphous immiscible blends if the
refractive indices are equal or if the size of the disperse phase is smaller than the

t.28

wavelength of ligh Therefore, turbidity is not sufﬁcient as the sole method for

determination of miscibility.

The method that is perhaps the most unequivocal criterion for miscible

blends is based on the glass transition temperature (Tg) determination. Phase-

178

separated blends have multiple Tgs, each characteristic of the respective

phases, while miscible blends have a single, composition dependent glass

transition temperature that falls at a temperature intermediate of those of the

individual homopolymers. Several empirical equations are available that relate T9

with composition. The most common is Fox’s relationship (Equation 7.7), which

can be used to predict the glass transition temperature of a miscible polymer

blend:
Tg Tgl TgZ

where wl is the mass fraction of the respective polymers and T9 is the glass

transition temperature.

Aside from thermal techniques such as differential scanning calorimetry,
differential thermal analysis, and dynamic mechanical analysis, electron
microscopy can also be used as the existence of multiple phases is a clear
indication of immiscibility.28 For practical measures of miscibility detection,

multiple methods should be used to ensure reliability.

Polymer Blend Morphology

The properties of immiscible polymer blends depend, to a large extent, on
the arrangement of the separate phases. In other words, the morphology of a
binary-phase blend governs the overall behavior. Two major morphologies can
be distinguished: isolated particles dispersed in a continuous phase matrix and a

two phase co-continuous morphology. In the former, the dispersed phase

179

typically corresponds to the lower mass fraction phase and will form isolated
droplets in the form of spheres, ﬁbrils, or platelets dispersed in a continuous
phase of the higher mass fraction polymer (the matrix). At higher mass fractions
of the dispersed polymer phase, the polymer droplets will coalesce and form
what is known as a co-continuous structure. In this morphology, two co-existing,
continuous and interconnected phases are present throughout the whole blend
volume.

In the case of having a dispersed phase within the matrix the properties of
the matrix phase will typically dominate. In contrast, in a co-continuous, two-
phase morphology the parallel arrangement allows for the properties of the blend
to be spatially isotropic; both components tend to contribute simultaneously to
the properties of the blend.”29

Additionally, the morphology of a polymer blend has dependence on a
number of parameters, including the polymer molecules, elasticity, interfacial
interaction, shear mixing, phase separation kinetics, and most importantly, the
melt viscosity of the individual components and the polymer component
composition.

White and Min30 studied the effect of viscosity ratio on the phase
morphology development during processing. The relationship between the melt

viscosity and volume ratio is schematically illustrated in Figure 7.2.

180

 

Continuous
phase of B

   
  
 

  
   
   

0"
conﬁnuous

Viscosity ratio A/B
I
I

Continuous
phase of A

 

 

I
Volume ratio A/B

 

Figure 7.2: Schematic representation of the effect of viscosity ratio and blend
composition on the morphology development of a polymer blend

From the ﬁgure, it can be deduced that the state of dispersion depends on
the concentration and the rheological properties of the individual components.
The higher proportion or the less viscous component forms the continuous
phase.31 The condition for dual phase co-continuity, as asserted by a study
authored by Jordhamo, Manson, and Sperling32, is the application of shear to a
polymer blend system close to the phase inversion region (the dotted horizontal
line in Figure 7.2). The model predicts that phase inversion should occur when
the viscosity ratio and the volume ratio are approximately equal i.e., when

171—=ﬂ— 7.8
772 ¢2 ( )

where 771 and 772 are the viscosities of phases 1 and 2, respectively, (61 and ¢2

are the respective volume fractions of phases 1 and 2.

181

Tailored Polymer Blend Properties

A property of a two component polymer mixture can be described by
Equation 7.9: 33

P = PICI + P2C2 + IPIPZ (7.9)

where P is a property value and P1 and P2 are the component property values.
C1 and C2 are the concentrations of the components and l is an interaction
coefﬁcient. The relationship between property and composition is illustrated in
Figure 7.3. The desired combination of properties results in I being equal to zero,
and the properties of the blend are a weighted arithmetic average of the
component properties. In most instances, blends are incompatible with [having a
negative value typically due to poor interfacial adhesion. In cases such as these,
the blend will display poorer properties than the additive result. Very seldom a
positive value of 1 generates a property synergy i.e. the resulting properties will

be better than the weighted mean of the individual components” properties. 33

182

 

Alloy I>0

 
 
   
 

Additive Result (l=0)

Property

Incompatible blend
(|<0)

 

 

 

0% 50% 100%
Concentration of Polymer 1

Figure 7.3: Property-composition relationship in polymer blends

Compatibilized Polymer Blends

As previously mentioned immiscibility dominates polymer blend
technology. Nevertheless, this does not mean that immiscible blends cannot
provide useful properties. Fortunately, immiscible blends can be compatibilized
and such blends cover a much broader scope of polymer mixtures than
domiscible polymer blends. Traditionally, compatibility is thought of as a polymer
mixture that does not exhibit any gross symptoms of phase separation. More
recently a more suitable deﬁnition was established to deﬁne a compatible system
as one which has one phase ﬁnely dispersed within the other phase, good
adhesion between blend phases, strong resistance to phase coalescence, and
possesses a desirable set of properties.34 Consideration of the latter deﬁnition
will be the basis of compatibility in the context of this work because oftentimes a

polymer blend with controlled two-phase morphology is desired because of its

183

ability to preserve the good features of each of the polymer components. In such
cases it is possible to still have improved properties by a synergistic effect of the
blend components. Compatibilization is essential for improved blend properties
usually by achieving the following:

1. Reduction of the interfacial tension, facilitating dispersion

2. Stabilization morphology against high stress and strain processing

3. Enhancement in adhesion between the phases in the solid state to

facilitate stress transfer

The above effects collectively result in a compatibilized product. Having a
disparity between surface energies of the respective components will yield poor
adhesion and may lead to phase separation and poor dispersion during
processing, resulting in brittle behavior. Thus, adhesion is a major issue and
must be fully understood in the development of polymer blends and composites
alike, as it plays a crucial role in governing the morphology and hence the
performance of the overall system.35 Suitably, many of the same principles that
apply to the composite adhesion theory discussed in Chapter 5 also apply here.
In both blends and composites the key is to develop techniques and processes
that will allow careful control over morphology and the interfacial polymer
components.

Compatibilization of binary polymer blends is very much inspired by
colloidial science. An example of this is the emulsiﬁcation of two Newtonian
liquids in which an emulsiﬁer (surfactant) is added to a hydrophobic/hydrophilic

mixture to improve the dispersion of one phase in another and to stabilize the

184

system. Polymer systems can be thought of in a similar way, where a
compatibilizer or coupling agent is analogous to the emulsiﬁer. The addition of a
compatibilizer improves the interfacial adhesion between the blend constituents.
In both types of systems the compatibilization effectiveness depends on the type
of equipment, mixing time, size of the dispersion, relative afﬁnity of compatibilizer
to the two polymers, molecular weight of the principal ingredients, the orientation
of the emulsiﬁer at the interface, and its ability to stabilize the interface against
ﬂocculation and coalescence. 36
In polymer blends the compatibilizer has segments that can speciﬁcally
interact with each of the main polymeric components. During mixing it will
preferentially locate at the interface between the phases and act by lowering the
interfacial tension between the components leading to a well dispersed
morphology. Various polymer systems are reported to have been compatibilized
by a variety of methods, yielding improved properties.‘°’7‘39 Some of such methods
include:40
o Incorporation of a core-shell copolymer that behaves like a multipurpose
compatibilizer
. Addition of block or graft copolymers (non-reactive blending) that are
miscible with one or both phases
0 Addition of bi-functional polymers with one part miscible with one
component and the second with the other component
. Induction of in situ grafting (reactive blending) designed to enhance

domain interactions

185

Reactive Compatibilization

Of speciﬁc interest in this project is the in situ formation of a copolymer
and its ability to improve the compatibility between immiscible pairs in polymer
blends. This commonly used and cost-effective method is arguably the best
strategy to produce new, multiphase polymeric materials from existing polymers
with superior properties. The copolymer formed at the interface is comparable to
a pre-made copolymer that is typically added at the interface of physical polymer
blends; it acts to reduce the interfacial tension between the phases, promote
adhesion between polymers, and stabilize the phase morphology?6 For the
successful execution of this technique four fundamental criteria must be
fufﬁlledz36'“

1. Adequate mixing to ensure the necessary distribution and dispersion of

one polymer within another

2. The presence of proper functional groups in each phase capable of

reacting across the polymer-polymer interface

3. The ability for reactions to occur in the extruder in the available

residence time

4. The formation of stable bonds across the interface during processing.

If these criteria are met, the result is a compatible blend whose degree of
compatibility is based on experimentally measured properties, some of which will

be discussed in the following section.

186

EXPERIMENTAL

Materials

PLA (4042D) was obtained from Natureworks LLC. PBAT (Ecoﬂex F
BX7011) (Figure 7.4) from BASF with a density of 1.26 g/cm3, a molecular weight
of 55,000 g/mol, and a melting temperature of 120°C was also used. Stannous 2-
ethylhexanoate (SnOctz) (Figure 7.5) was purchased from Sigma Aldrich and

used as the transesteriﬁcation catalyst.

O O
4“ Q NH er II ICI
O—C O—(CH2)4 O—C—(CH2)4—

Figure 7.4: Structure of poly (butylene adipate co-terephthalate) (PBAT)

0
/\/\ﬁ‘\O-Sn++
0%
Figure 7.5: Structure of stannous 2-ethylhexanoate (SHOCtz)

Methods

Reactive Blending of PLA/PBAT Blends
Prior to extrusion, PLA and PBAT pellets were oven dried at 85°C for 24

hours. PBAT, PLA, and catalyst were weighed and manually tumble-mixed
before being fed into the feeder of a Century ZSK-30 twin screw extruder. The
compositions of the blends are listed in Table 7.1. The extruder barrel is

comprised of nine zones, whose temperatures are presented in Table 7.2. All

187

polymer blend compositions were extruded at 150 rpm. Upon exiting the die the
extrudate was cooled through a waterbath and pelletized downstream. Pellets

were collected and oven dried for 48 hours at 65°C.

Table 7.1: PBAT/PLA Blend Compositions

 

 

 

 

Blend Composition (wt%)
PLA PBAT SnOctz
30 70 O
50 50 O
70 30 O
30 70 0.05
50 50 0.05
70 30 0.05

 

30 70 0.10
50 50 0.10
70 30 0.10
30 70 0.50
50 50 0.50
70 30 0.50

 

Table 7.2: Extrusion Processing Conditions for PBAT/PLA Blends

 

Temperature (°C)
Zone 1 2 3 4 5 6 7 8 9 Die
25 95 160 165 170 180 180 180 170 165

Blown Film Extrusion
The dried extrusion-blended pellets were blown into ﬁlms using a Killion

single-screw blown ﬁlm extruder with a screw diameter of 25.4 mm and a length-
diameter ratio of 251. Films were blown through an inner die with a diameter of
50.8 mm and a die gap size of 1.5 mm. The ﬁlm blowing conditions are shown in

Table 7.3.

188

Table 7.3: Blown Film Processing Conditions for PBAT/PLA Blends

 

 

Temperature QB)
Zo1ne 202% Kane Clamp Ring Adaptor Die 1 Die 2
205 205 205 205 ' 202 199 174

 

Mechanical Properties
Tensile tests were performed using a UTS Mechanical Testing Apparatus

(Model SFM-20) ﬁtted with a 100 lb load cell. Testing was done as per the ASTM
D88243 standard on specimens mounted both parallel to the axis of ﬁlm
processing (the machine direction) (MD) and perpendicular to the processing
axis (the transverse direction) (TD), with the average of 8 specimens reported.
Tear resistance was determined using the trouser test method (ASTM D 1938-
02a)44 on ﬁlm specimens apprOximater 1 mm in thickness.

Differential Scanning Calorimetry (DSC)

Thermal transitions of the blends were measured by means of DSC. Film
samples were heated to 200°C at a rate of 30°C/min and held for 5 minutes in a
nitrogen atmosphere. With prior thermal history erased the samples were cooled
at a rate of 30°C/min then reheated to 200°C at a rate of 5°C/min. Cold
crystallization, melting, and glass transition temperatures were obtained from the
second heating scan. The degrees of crystallinity of the composites were
calculated according to Equation 3.2, using 115 J/g or 93.7 J/g for the respective
heats of fusion for PBAT and PLA.

Thermogravimetric Analysis (TGA)

A high resolution thermal gravimetric analyzer (Model 2950) from TA

Instruments was used to determine the thermal decomposition temperature of

189

the neat polymers and of the blends. Films were heated from room temperature
to 500°C at a heating rate of 10°C/min under a nitrogen atmosphere. The thermal
stabilities of the blends were evaluated based on the weight loss percentage
versus temperature curves.
Thermo-mechanical Analysis

Dynamic mechanical analysis (DMA) was performed on a TA 2980 DMA.
Film specimens were cut along the machine direction having dimensions of 30
mm (length) x 6.45 mm (width) x 0.1 mm (thickness) and mounted in a tension
clamp. All ﬁlms, except for the PLA cast ﬁlm, were heated from -50°C to 140°C at
a rate of 2°C/min. The PLA ﬁlm was heated from 20°C to 140°C. Tests were run

at an amplitude of 50 pm and a frequency of 1 Hz. Storage modulus (E’), loss

modulus (E"), and tan 8 (E”/E”) were recorded as a function of temperature.

Gel Permeation Chromatography (GPC)
Molecular weight measurements of the blends were carried out in a

Waters GPC system equipped with an isocratic 1515 HPLC pump (ﬂow rate = 1
mUmin), a 2414 refractive index detector, and a 717 autosampler. Samples were
dissolved in tetrahydrofuran in concentrations of 2 mg/mL. Molecular weights (Mn
and MW) and molecular weight distributions were calculated using a universal
calibration curve constructed with polystyrene standards.
Transmission Electron Microscopy (TEM)

The effects of catalyst concentration on the morphology of the blends
were examined using TEM. Using an ultramicrotome, samples were

cryogenically sectioned from an extruded pellet in a plane perpendicular to the

190

extrusion direction. For each sample, a forrnvar ﬁlm was used for support and
mounted on a copper grid. Sections were examined under a JEOL JEM-100CX II
TEM through condenser aperture #1, using objective aperature #2, and an
accelerating voltage of 100 kV.
Statistical Analysis

Statistical analysis was performed using SPSS software for Windows.
Data were compared using analysis of variance (ANOVA) and a Tukey post-

ANOVA test (p _<_ 0.05).
RESULTS AND DISCUSSION

Proposed Transesteriﬁcation Reaction

A possible transesteriﬁcation reaction between the PBAT and PLA is
depicted in Figure 7.6. The proposed reaction induces the in situ formation of a
diblock copolymer at the interface between the two phases in the phase-
separated system. The copolymer is expected to act as an interfacial modiﬁer to
strengthen the interfacial region between the blend components resulting in

improved properties of the blend.

191

 

H3
l
+O—C CHO—(CH2)4}P‘C—(CH2)4—IC++ +‘[':: C—CI:+
PBAT
A
SHOCtz
“1' ‘1? l? ‘1’ 1' 9 (u)
O—C—C—ﬂCth-"O—C—(Cth—C‘} OO—C—C—(CH2)4—C+
CH3 CH3
‘1’ ﬁ’ 9'3 1??“3
O—C C-O—(‘Z O—C C—O—(CH2)4- O—C- (‘3
H H

 

 

 

Figure 7.6: Possible transesteriﬁcation reaction between PLA and PBAT

Mechanical Properties of Blends

Tensile Behavior
The tensile properties of the ﬁlms are given in Figures 7.7 and 7.8. In the

physical blends increasing the PLA content was seen to progressively increase
the tensile strength and decrease the elongation in both the machine and
transverse directions. This was expected because PLA is stronger and more
brittle than PBAT. Upon the addition of catalyst, a noticeable effect was seen in
the low PLA composition. In these samples, the tensile strengths of the blends
containing a minimal amount of catalyst were enhanced compared to the
physical blend. For example, in the blends containing 30% PLA, the addition of

0.05% catalyst gave the most signiﬁcant increase from 31.3 MPa to 58.4 MPa in

192

the machine direction and 15.3 MPa to 35.3 MPa in the transverse direction. At
higher PLA concentrations, however, the addition of catalyst resulted in no
signiﬁcant change in strength as compared to the physical blends. It was also
apparent that the tensile properties began to deteriorate at catalyst levels
exceeding this optimum concentration and this was the case for ﬁlms tested in

both the machine and transverse directions.

 

 

Tensile Strength (MPa)

 

 

 

0 30 PLA (%) 50 70

F Ino catalyst added I0.05% catalyst D0.1% catalyst D0.5°/o catalyst ]

 

 

Figure 7.7: Tensile strength of PBAT/PLA ﬁlms tested in the machine direction

 

O)
O

 

-| N (.0 h 0!
O O O O O
I A I L

Tensile strength (MPa)

     

 

 

O
1

 

o 30 PLA (wt %) 50 70
I Inc catalyst added l0.05% catalyst 130.1% catalyst 130.5% catalyst I

 

 

 

 

193

Figure 7.8: Tensile strength of PBAT/PLA ﬁlms tested in the transverse direction

The effect of catalyst concentration on the elongation was also examined.
Figures 7.9 and 7.10 show that at a low PLA concentration (30%), the SnOctz
addition resulted in a continuous increase in elongation at break with catalyst in
both directions, but particularly in the machine direction. The sample containing
50% PLA showed an increase only with 0.05% SnOctz. Beyond this catalyst
concentration a drastic decrease in elongation at break was observed. The
samples containing 70% PLA displayed brittle behavior, characteristic of neat
PLA, which further decreased with catalyst addition. It is known that the
elongation at break of a blend is sensitive to the strength at the interface between
components, reﬂecting the degree of compatibilization. Thus, the improved
elongation is an indication of a stronger bond between components. The lack of
improvement at high PLA concentrations suggests minimal mixing in this
composition range; otherwise, the PBAT phase would have had a toughening

effect on the PLA rather than increasing the brittleness.

194

 

 

 

 

 

 

o 30 50 7o
PLA (%)

I I no catalyst added I 0.05% catalyst D 0.1% catalyst E 0.5% catalyst I

 

 

 

 

(D
C)
O

 

on
000°
000°

00
OO

Elongation at break (%)
—h N 0.) #- 01 O) \I
O O
O O

 

 

 

. L

Figure 7.9: Elongation at break of PBAT/PLA ﬁlms tested in the machine
30 50 70

direction
0 I ,
0 PLA (%)

I Inc catalyst added l0.05% catalyst 00.1% catalyst D0.5% catalyst I

 

 

 

Figure 7.10: Elongation at break of PBAT/PLA ﬁlms tested in the transverse
direction

195

 

Not only are the mechanical properties of the ﬁlms dependent on the
intermolecular forces, chain stiffness, and molecular symmetry of the individual
polymers, they also depend on the orientation of the ﬁlms. It was observed that
the mechanical properties of the ﬁlms tested in the machine direction were.
slightly higher than those tested in the transverse direction. Blown ﬁlms are
typically anisotropically oriented, with their molecular chain axis lying in the plane
parallel to the processing direction. The alignment due to the blown ﬁlm
processing dominated the ﬁlm’s properties and can be explained by realizing
that, parallel to the orientation (in the machine direction), stresses are exerted
largely on the primary bonds of the polymer chain, while in the transverse
direction, forces are applied to the weaker secondary bonds between polymer
chains.45

Orientation also played a role in the elongation at break of the blends. The
samples containing a low PLA content showed greater elongations at break in
the direction perpendicular to the direction of orientation, which is the opposite of
what was observed in the tensile behavior. This can be best explained by the re-
orientation that takes place among the molecules that occurs as a perpendicular
stress is applied. The molecules ﬁrst disorient; then, when stress is applied they
re—orient themselves in the direction of the applied stress. This, consequently
results in an elongation at break characteristic of an elongation ensuing from the
parallel direction, combined with that associated with the disorientation and re-

orientation of the molecules. 45

196

Tear Strength
Plots of tear resistance versus blend composition for ﬁlms tested in both

the machine direction (MD) and the transverse direction (TD) are shown in Figure
7.11 and Figure 7.12, respectively. In both directions the tear resistance of the
blends increased with increasing SnOctz content for ﬁlms with 30% PLA.
However, the overall tear strength in the TD was greater than the corresponding
tear strength in the MD. Cleariy, this behavior is plausible due to the formation of
chemical or physical linkages between the two polyesters. These linkages may
have resulted in a network structure and thus, induced increased tear strength in
the transverse direction all while restricting chain movement in the longitudinal
direction."'6

Furthermore, the superior tear strength of the catalyzed blends can be
attributed to increased compatibilization. Li and Wong47 showed that, in general,
tear resistance increases as the degree of transesteriﬁcation increases, but
decreases with a decline in molecular weight. Therefore, optimal tear strength
may be achieved with conditions that favor a balance between a high degree of

transesteriﬁcation and minimal degradation.47

197

 

 

.N
U‘I

N

3"
U1

 

 

 

Tear Resistance (Kg)

 

 

 

0.5
o I
30 50 70 I
PLA (wt%) I
[Inc catalyst added I 0.05 wt% SnOct2 00.1 wt% SnOct2 '05 wt% SnOct2I .;

 

Figure 7.11: Tear resistance of PBAT/PLA ﬁlms tested in the machine direction

 

 

 

 

 

 

’5
1‘,
d)
0
C
N
'63
"a
Q,
I!
'33
I-
30 50 7o
PLA (wt%)

 

 

I no catalyst added l0.05wt% SnOct2 no.1 wt% SnOct2 1110.5 wt% SnOct2I

 

 

Figure 7.12: Tear resistance of PBAT/PLA ﬁlms tested in the transverse
direction

It is important to note that with regard to the mechanical properties, the
effects of transesteriﬁcation are dependent upon the composition of the blend. It

relies on whether PBAT or PLA makes up the major phase. When PLA is the

198

minor component, the tensile strength, tear strength, and elongation at break
values of the blends prepared with catalyst show improved properties over those
of the physical blends. This is because this composition is more suitable for
resistance to failure due to the blend morphology, which will be discussed in a
subsequent section. Conversely, for the blends in which PLA was the major
component, the catalyst addition had an adverse effect on the properties. This
could be due to the catalyst-induced degradation of PLA or to poor mixing during
the reaction, causing dramatic embrittlement and lower tear properties. On the
basis of mechanical property analysis, there appears to be more association

between the polymer phases when PBAT was the majority phase.

Molecular Weight of PBA T/PLA Blends

Molecular weight is an important factor in reactive blending because the
mechanical properties of the resulting blends depend largely on the molecular
weight of the polymer components. To understand the effect of blending and
catalyst concentration on the PBAT/PLA blends, the molecular weights of the
neat polymers along with the PBAT/PLA blend compositions with and without
catalyst were determined by GPC. The chromatograms are presented in Figure
7.13. Each of the 30:70 PBAT/PLA (Figure 7.13a) blends displayed in multiple
peaks, where the 70:30 PBAT/PLA (Figure 7.13b) and the 50:50 PBAT/PLA
(Figure 7.130) blends each exhibited one peak indicating the compatibilization
effect of the catalyst at higher PBAT concentrations.

In the 30:70 PBAT/PLA blends, both peaks broadened and decreased in

intensity with increasing catalyst concentration, showing predominant

199

degradation of the PLA phase. The other blend compositions also experienced a
shift to higher retention times with increased catalyst concentration indicating
lower molecular weights.

The GPC data of the 70:30 PBAT/PLA samples are summarized in Table
7.4. As shown, the molecular weights decreased with SnOct2 concentration.
From the mechanical property data it is assumed that the SnOct2 promoted
transesteliﬁcation between PBAT and PLA. However, it has been reported that in
excess transesteriﬁcation reactions in PLA are associated with reverse
depolymerization, leading to lower molecular weights, broadening of molecular
weight distribution, and deterioration of polymer properties.“

These data, along with investigation mechanical properties, suggest that
transesteriﬁcation promoted degradation and copolymer formation, both of which
contributed to the properties of the blends. When the SnOct2 content was low
(0.05%), copolymer formation dominated the properties, evidenced by increased
tensile properties. However, with higher SnOct2 content the properties were
governed by the degradation, explaining the lower molecular weight and

decreased strength.

200

 

my; - . -.4—

1,.

 

 

0-5 — — PBAT/PLA (30 70) I (a)
—— PBAT/PLAlSnOth (30: 70: 0 05) I \

    

0.4 \

----- PBAT/PLAlSnOth (30 70: 0 1) I (
0,3 — -- PBAT/PLA/SnOth (30:70:05) I, I.""\\\
0-2 \
0.1

    

 

  

 

   
 
 
  

0 ..... .., ...-~ '12,; 1'1???
"""”TU""”": 12 ""73 16 18 20 "'2’27"
41.1 I
Retention time (min)
(b)
0-5 — -PBAT/PLA (70:30)
-—- PBAT/PLAlSnOth (70:30:005)
0.4 ----- PBATIPLAISnOth (70: 30: 0 1) II; x '\
PBAT/PLNSnOth (70:30: 0 5) ‘-
0.3
02

 

 

(C)
— — PBAT/PLA (50:50)
04 -——- PBAT/PLAISnOct2 (50:50:00.5) at}.
----- PBAT/PLAISnOct2(50:50:0.1) ‘
0.3 — - - PBAT/PLA/SnOct2 (50:50:05)
02
0.1

 

 

 

--. 101-Puf- 1 8 20 22
-0.1 Retention time 6(min)

 

 

Figure 7.13: GPC traces of PBAT/PLA blends for the following compositions:
(a) 30:70; (b) 70:30; (0) 50:50

201

Table 7.4: Molecular Weights and Molecular Weight Distributions of PBAT/PLA

 

 

 

Blends
Blend
Composition Mn (g/mol) Mw (9/mol) MW/Mn
PBAT 80026 i 504 114796 i 693 1.43 :l: 0.02
PLA 105488 1: 524 166156 1: 497 1.58 :1: 0.01
PBAT/PLA
70/30 85570 i 1324 126477 1 1839 1.48 d: 0.05
PBAT/PLA/SnOctz
(70/30/0.05) 72505 i 1577 113450 :1: 2737 1.56 d: 0.07
PBAT/PLA/SnOctz
(70/30/0.1) 64164 1: 1797 105845 :1: 2577 1.65 i 0.08
PBAT/PLA/SnOctz
(70/30/0.5) 55772 :t 1185 95970 :1: 1689 1.72 1: 0.07
Thermal Properties

Differential Scanning Calorimetry
Within one polymer the presence of another polymer can, in effect, alter

the cooperative segmental movement of the polymer chain, changing its thermal
transition behavior. Therefore, the determination of the thermal transitions of
blends proves to be useful in revealing interactions between respective polymers.
Table 7.5 shows the transition temperatures of the physical blends and of the
70:30 PBAT/PLA compositions with varying catalyst concentrations. PBAT and
PLA each showed one glass transition (T9) identiﬁed at 597°C and -32.5°C,
respectively. The blends were characterized by two distinct Tgs, representative of
their parent polymers. The T95 of the PBAT component in the blends shifted to
higher temperatures, but were essentially independent of the nominal

composition, whereas the PLA component showed minimal change. This

202

represents the lack of signiﬁcant molecular interactions between the polymers,
conﬁrming the presence of separate phases after the initial melting.49

In the catalyzed blends, the addition of catalyst appeared to have little
effect on the thermal behavior at all blend compositions. Like the neat blends, the
catalyzed blends also showed two distinct Tgs, corresponding to the separate
polymers, conﬁrming the immiscibility of the polymers, even with added catalyst.
At high catalyst concentration, the T9 of the PBAT component shifted from -9.9°C
in the physical blend to -6.1°C in the blend containing 0.5% catalyst, while that of
the PLA phase shifted from 576°C to 488°C. Since the T9 is a measure of
molecular mobility of the chains, these shifts, although minimal, were attributed
to active interaction between PLA and PBAT chains at 0.5% catalyst. The fact
that the PLA component shifted by a larger degree shows that the
transesteriﬁcation reaction has more of an effect on the PLA phase than on the
PBAT phase.

The temperatures of cold crystallization (Tc) shifted to lower temperatures
and exhibited narrower peak widths, indicating an enhanced crystallizability of
PLA in the blends containing catalyst. This shift in TO parallels the convergence in
glass transition temperatures and points to the possibility of enhanced
compatibility in this temperature regime as a result of high catalyst addition.

The melting behavior of the physical blends was characterized by an
endotherm and a small shoulder. In the reactive blends, the addition of catalyst
appeared to separate the melting endotherm and shoulder into two distinct

endotherms. While multiple endotherms are quite common in PLA-based

203

polymers 50' 51, their presence quite possibly could indicate the formation of a new
crystalline structure induced by the coupling of PLA and PBAT. The lower
temperature endotherm shifted to lower temperatures while the higher one
remained unchanged with increasing catalyst. The decreased melting
temperature is a result of defects in the crystallites resulting from the presence of
PLA/PBAT copolymers or other by-products of the interchange reactions.

The degrees of crystallization of the blends, 10 are also shown in Table
7.5. Of the physical blends, that which contains 70% PBAT possessed the
highest crystallinity (52.7%). This value decreased as PLA content increased. In
the reactive blends, the crystallinity initially decreased with catalyst concentration
to 11.2%, then increased to 23% at 0.5% catalyst. The increased crystallinity at
the high catalyst concentration can be related to the lower molecular weight of
this system. Within a lower molecular weight polymer, polymer chains can easily

rearrange and align themselves to result in a more crystalline structure.

204

ﬂ

1;- .

._o>m_ cocouccoo oxoma 05 am 80:98:“. Ewescaﬁ E3659 5:28 0:56 65 55:5 $652 E2920

 

 

 

 

umd H 9mm cwd H a.aa do... H ma: cod H Naar 1 and H adv umd H mm. N manomnon
HoocmSmPaamn.

cad H N. E. and H a. F E. and H mamr G__..a H :3. 1 mm... H «.mm owe H Van. N monomuon
HoOCmSaPaqma

and H it one H odor mNéH «ta: avd H ad: 1 a: H mam 85.0 H fa- N madam”?
aoOcm\<l_E._.<mn.
6258 5? mcoEmanoO
cud H mom and H ad: «Yo H ma: 1 -- mad H 98 ova H 0N- ohom Sap/«mm
and H new nod H ...m: 50.0 H ma: 1 1 and H mam aﬁo H ma- omuom <._E._.<mn_
dad H Nam avd H «6:. mad H «a: -- 1 «No H 0.5m and H a.a- canon <._n=._.<mn_
sud H ow... DEN H ma: 1 mod H ad? I mwd H Nam I (...n.
mod Homo a: H ads 1 I E. H tom? 1 No.0 HmNm- ._.<mn_
@1628 305.3. muoEmOQEoo

o
as a ...N. a” 2a.. ...e. as...

 

855 592?. 5 £8 08 ”m: was.

206

Thennomechanical Analysis

The dynamic mechanical response of a material is sensitive to the quality
of bonding between polymers. Therefore, as a complementary tool to DSC, DMA
was used to investigate relaxation phenomena in the blends. The dynamic
mechanical properties of the physical PBAT/PLA blends, along with those of the
pure polymers, are shown in Figures 7.14 through 7.16. The representative
curves of the dynamic storage moduli as a function of temperature are depicted
in Figure 7.14. It is apparent that PBAT and PLA each exhibit sharp, one-step
glass-to-rubbery transitions whereas the blends show intermediate broad two-
step transitions.

The blends showed a progressive increase in storage moduli as the PLA
content increased. Increasing the PLA content suitably decreased the ﬂexibility of
the blend and correspondingly increased the dynamic storage moduli over that of

pure PBAT.

206

 

 

 

 

 

 

  

 

 

 

1
1 o PBAT
, :1 PBAT/PLA (70:30)
. + PBAT/PLA (50:50)
’3 ‘ o PBAT/PLA (30:70)
0_ I PLA
2 _
V
m
2 .
3 l
U 1
o ‘l
2 .
a)
CD-
‘3
o
H q
(I) . ., g
- \N
1..
-60 -10 140

40
Temperature (°C)

Figure 7.14: Storage modulus curves of PBAT/PLA physical blends

Figure 7.15 shows the loss modulus curves of the blends and of the pure
polymers. The maximum peak intensity of the loss modulus curves is directly
proportional to the volume fraction of the individual phases present in the blend.
Increasing the PLA concentration in the blend resulted in a corresponding
increase in peak intensity. In addition, the temperatures corresponding to the
peaks represent the Tgs of the polymers. PBAT and PLA have respective peaks
at -22°C and 71°C. Their blends had two peaks, each corresponding to the PBAT

and PLA components.

207

 

 

O PBAT

1:1 PBAT/PLA (70:30)
+ PBAT/PLA (50:50)
0 PBAT/PLA (30:70)
I PLA

 

 

 

33':th

Loss Modulus (MPa)
“F—I-r—t—I—uIm—g—g
__.—+|-—|—-l-—H

 

 

 

 

-60' '-1'0 _ 40 90 ' 140
Temperature (°C)

Figure 7.15: Loss modulus curves of PBAT/PLA physical blends

Another parameter related to the glass transition is the loss factor (tan 6),
which is shown as a function of temperature for the PBAT/PLA blends in Figure
7.16. It represents the changes taking place in the polymer molecules during
dynamic mechanical testing. Similar behavior to the loss moduli was seen in
which no major shifts in the lower temperature transition was discernible, while
the peak corresponding to PLA gradually shifted to lower temperatures with

decreasing PLA contents.

208

 

 

 

o PBAT

El PBAT/PLA (70:30)
+ PBAT/PLA (50:50)
0 PBAT/PLA (30:70)
I PLA

 

 

 

Tan Delta

 

I

 

 

 

 

 

-60 _ -10

40
Temperature (°C)

Figure 7.16: Tan 8 curves of PBAT/PLA physical blends

The storage modulus curves of the 70:30 PBAT/PLA catalyzed blends are
shown in Figure 7.17. Similar to the physical blends, two relaxation processes
can be distinctively identiﬁed and are associated with the transition of each
phase. The shapes of the curves were identical; however, the magnitudes of the
curves changed with catalyst concentration. The magnitude and nature of the
change in the storage modulus are dependent on the overall composition of the
polymer blends and are determined both by the intermolecular and
intramolecular interactions; the latter having greater inﬂuence in the different
physical states of the polymer.52 The low temperature inﬂections that correspond

to the relaxation of PBAT are apparent between -30°C and -10°C, while the

209

relaxations around 50°C are attributed to the glass-to-rubber transition of PLA.
The blend prepared with 0.05% catalyst displayed the highest storage modulus,
followed by the sample prepared with 0.5% catalyst, the physical blend, and
ﬁnally the blend with 0.1% catalyst. The concentrations of PLA and PBAT in the
blends were essentially constant throughout all of the samples, indicating that
any differences in moduli are a result of differing morphologies caused varying
catalyst concentrations. Going from completely incompatible phases in the
physical blend to a more compatible blend, the modulus increased due to an
increase in interfacial interaction resulting from the transesterﬁcation reaction. An
exception is the sample prepared with 0.1% catalyst, which displayed the lowest
storage modulus values across the entire temperature range. This can be
explained by the morphological differences in this sample and will be discussed

in the following section.

210

 

 

 

 

 

 

‘ O PBAT/PLA (70:30)
, c = :1 PBAT/PLA/SnOc (69.97:29.99:0.05)
. 5 .. + PBAT/PLA/SnOc(69.93:29.97:0.1)
11?. .. o PBAT/PLA/SnOc (69.65:29.85:0.5)
o. . ..
é . II
w - II
2 LI
3 ' ll
'0 l -‘
O .1
2 ‘
<0 -
O) 4
E .
O
,_. .
(D -
I T

 

 

 

-60 ' ' ' -10 ' 140

Temper‘agure (°C)
Figure 7.17: Storage modulus curves of PBAT/PLA curves with varying catalyst
concentrations

The loss modului of the catalyzed blends are shown in Figure 7.18, in
which the peak positions represent the temperatures associated with the glass
transitions. The Tgs representative of the PLA component were found to slightly
shift to lower temperatures, which indicate some interaction between PBAT and
PLA with catalyst addition. These temperatures, as well as those of the physical
blends, differ signiﬁcantly from the Tg results found by DSC. DSC and DMA have
different sensitivities as a result of testing in either static or dynamic modes. DMA
is usually more sensitive for identifying transitions in polymeric materials, such as
the additional transition that appears at 100°C in the 0.5% catalyst system that is

not apparent in its DSC thermogram.

211

The additional peak, interestingly, was not apparent in any of the other
blend compositions. Extra peaks in loss moduli curves of various systems have
been reported and result from a variety of different pheonomena.°3"54 Similarly,
the tan 6 plots (Figure 7.19) yielded two additional peaks located at around 80°C
and 100°C. These additional peaks are attributable to either: (1) the glass
transition temperature of the new copolymer formed at the interface as a result of
the transesteriﬁcation reaction; (2) the recrystallization of PLA; or (3) the

formation of thermal lag in different portions of the specimen. 53' 54' 55

 

 

 
 
  
 

 

 

 

 

 
 

 

 

j o PBAT/PLA (70:30)
, .1 1:1 PBAT/PLA/SnOc (69.97:29.99:0.05)

_ + PBAT/PLA/SnOc (69.93:29.97:0.1)
A o PBAT/PLA/SnOc (69.65:29.85:0.5)
m l
n. .
é .
m -
2
§ .
E -
a) .
(D
o .
_l

-60'°'-10r'7 9 140

40 '
Temperature (°C)

Figure 7.18: Loss modulus curves of PBAT/PLA curves with varying catalyst
concentrations

212

 

 

O PBAT/PLA (70:30)
a PBAT/PLA/SnOc (69.97:29.99:0.05)
+ PBAT/PLA/SnOc (69.97:29.93:0.1)
o PBAT/PLA/SnOc (69.65:29.85:0.5)

 

 

 

Tan Delta

 

 

 

 

Figure 7.19: Tan 8 curves of PBAT/PLA curves with varying catalyst
concentrations

Thermogravimetry
TGA was performed on the samples and the weight loss, due to the

volatilization of the degradation products, was monitored as a function of
temperature. Figure 7.20 reveals the degradation behavior of the neat polymers
and the physical blends. PLA and PBAT each exhibit one-step degradation
proﬁles and the blends show multiple-step degradation processes with varying
rates of degradation. In the initial stages of degradation the thermal stabilities of
the blends increased as PBAT concentration increased and were intermediate to
the two polymers. The reverse was true in the later stages of degradation, where

the thermal stabilities of the blends exceeded those of PBAT.

213

 

 

 

 

 

 

 

 

 

120
o PBAT
100‘ 1:1 PBAT/PLA (30:70)
. .. + PBAT/PLA (50:50)
. ‘1 o PBAT/PLA (70:30)
1 u I PLA
80: I
5" '
560‘
g ‘ ‘\
40- \l
l \-
F .\ t
20 “1..
\ \&
0 . . . , —. . — — ———- —=
225 325 425

Temperature (°C)

Figure 7.20: TGA thermograms of PBAT/PLA physical blends

TGA proﬁles of the 70:30 PBAT/PLA catalyzed blends are displayed in
Figure 7.21. Each curve shows decreased thermal stability with catalyst addition.
To quantify the effect of added catalyst on the thermal degradation of the blends
differential thennogravimetry (DTG) proﬁles were studied. From these ﬁrst
derivative curves (Figure 7.22) the decomposition temperatures (Tmax) were
determined, which represent the maximum rate of weight loss. Each blend
exhibits a two-step decomposition proﬁle, except for the blend prepared with
0.5% catalyst. This blend underwent a three-step decomposition conﬁrmed by

the presence of a third peak. The additional Tm...x peak was ascribed to the Tmx of

214

the copolymer that was formed or to that of the degraded polymer, which was

possibly facilitated by the thermal degradation products of PLA.

 

 

 
 
   

 

 

 

 

 

 

120
I o PBAT/PLA (70:30)
1 :1 PBAT/PLA/SnOc (69.97:29.99:0.05)
100 + PBAT/PLA/SnOc (69.93:29.97:0.1)
: I o PBAT/PLA/SnOc (69.65:29.85:0.5)
804 1’
(a 1 b.
e,
560
a) .
g .
40-
20-
325 ' Y ' 325 ' ' ' 425

Temperature (°C)

Figure 7.21: TGA thermograms of PBAT/PLA blends showing the effect of
catalyst concentration on the decomposition behavior

215

 

 

 

 

 

 

 

 

o PBAT/PLA (70:30) .0
n PBAT/PLA/SnOc (59972999005) g
+ PBAT/PLA/SnOc (69.93:29.97:0.1) +1,
OPBAT/PLA/SnOc(696522985205) - .. :9
a It PH.
8 HH-O—Q-O—O-O—OHO—G-o-O-O ‘9’“ KOH“
“ 't
in 1L4, \
é) HW'Hﬁ—H-t-H-‘H-‘r-H n lg, +—++—+-+-
>. Ilia 1:11
'5 31:1,]. 3 Ink!
0 aa-a—a—aaa—ana—a-aaa—a-aa—a-a ‘ﬁﬁai ‘B-aaa—a-
0Y'Y100' ' '200' '400' ' '500

Temperature3(9(()3)

Figure 7.22: Differential thermogravimetric curves of 70:30 PBAT/PLA ﬁlms

showing effect of catalyst concentration on the decomposition temperatures
Table 7.6 lists the onset temperature (Tm) and the Tm.x values of the
catalyzed blends. The Tm decreased upon catalyst addition with no signiﬁcant
changes thereafter (p 2 0.05). The Tmax values corresponding to PLA decreased
with catalyst concentration, while those corresponding to PBAT remained
unchanged until a catalyst concentration of 0.5% was reached, at which time the
Tmx decreased by nearly 30°C. These results indicate that SnOct2 promoted
thermal degradation in PLA more so than in PBAT. The structural changes that
resulted from the compatibilization reaction interfered with the thermal stability of
the parent polymers, particularly PLA. Such changes apparently resulted in a

structure that was more susceptible to thermal degradation.

216

Table 7.6: Thermogravimetric Parameters of PBAT/PLA Blends

 

 

Sample T1% (°C) Tmax (°C) Tmax (°C) Tmax (°C)
PBAT 324716.38 368.8:353 -- --
PLA 316412.13 330.4101b - --

PBAT/PLA 70:30 314.9339a 339.6137b 368922.68 --

PBAT/PLA/SnOctz b c 8
70330305 292.8:33 305.7151 369.9149 .-
PBAT/PLA/SnOctz b c a
703301,.1 291.8:41 301.3:28 369.9106 --
PBAEQOLI’EIOSQOQZ 277.3171b 285.3112d 340.316.1b 373.8127

 

Different letters within the same column represent signiﬁcant differences at the 95% conﬁdence
level.

Blend Morphology

The majority of polymer blend properties are inﬂuenced by the ﬁnal
morphology. Therefore, in studying the compatibilization of polymer blends, it is
necessary to understand the relationship between polymer morphology and the
properties it dictates. The morphology of the PBAT/PLA blends was analyzed by
examination of the cryo-sectioned surfaces of the blend pellets. Contrast was
apparent without staining and evolved from the structural differences between
the polymers. PLA is considered “electron transparent” because inelastically
scattered electrons are transmitted through the objective lens aperture, rather
than being intercepted by it.°° Because of this, PLA appears as the lighter phase.
The PBAT phase, being more electron-dense, was able to scatter and absorb

more electrons, therefore appears as the darker phase.

217

 

Each blend micrograph depicts a binary structure with one or more of the
following phase morphologies:
(1) Dispersions of one phase containing a portion of the continuous
phase to form composite inclusions
(2) An interpenetrating co-continuous structure in which both phases
of equal composition co-exist in an interconnected network
structure
(3) A morphology in which droplets of one phase are dispersed
throughout a continuous phase of the other polymer
Examples of each of these phase morphologies are shown in Figure 7.23
through Figure 31. In the PBAT/PLA 70:30 physical blend (Figure 7.23), PBAT is
the continuous phase containing PLA island-like domains. Within the PLA phase
are PBAT sub-inclusions. This morphology, whereby the dispersed phase
contains small domains of the matrix phase, represents a phenomenon known as
phase inversion and has been studied by numerous researchers.”59 This
occurrence is believed to be associated with a thermodynamic change where the
dispersed phase inverts, becoming the major continuous phase in order to
minimize the free energy.°° Phase inversion is related to the viscosity ratio and
proper control of mixing. It is postulated that the lower viscosity PLA initially
formed the matrix containing droplets of PBAT. As the PBAT concentration
increased and began to disperse in the PLA component matrix, a percolation
threshold was reached, causing the blend to invert in phase continuity. Similar

morphology has been reported in studies by Favis and Chalifoux‘51 on PE/PS and

218

PP/PC blends and Astruc and Navard62 on PDMS/HPC (hydroxypropylcellulose)
blends.

As the PLA content increased, the size of its domains also increased and
began to form a semi-continuous phase in which PBAT was dispersed, while
PBAT also existed as the continuous phase (Figure 7.24). In this PBAT/PLA
50:50 blend the morphology is considered to be co-continuous. With more PLA,
as shown in Figure 7.25, PLA became the continuous phase as a result of the
higher concentration and lower viscosity of PLA, both of which promote droplet

agglomeration during extrusion processing.22

 

Figure 7.23: TEM micrograph of the 70:30 PBAT/PLA physical blend

219

 

Figure 7.24: TEM micrograph of 50:50 PBAT/PLA physical blend

220

 

Figure 7.25: TEM micrograph of 30:70 PBAT/PLA physical blend

Effect of Catalyst on the Morphology

TEM micrographs representative of the blends prepared with 0.05%
SnOct2 are shown in Figures 7.26 through 7.28. A dramatic morphology change
was observed from that of the physical blends. Comparison of the catalyzed and
uncatalyzed blends shows that the morphologies transformed from having larger
phases to being composed of much smaller, discrete phases. This change is
particularly apparent in the 70:30 PBAT/PLA and 50:50 PBAT/PLA blends.
Speciﬁcally, in the catalyzed 70:30 PBAT/PLA blend (Figure 7.26) the domain

size of PLA was signiﬁcantly reduced as compared to the uncatalyzed blend in

221

Figure 7.23. In the catalyzed blend well deﬁned particles were dispersed
throughout the PBAT matrix and elongated parallel to the extrusion direction. In
the intermediate composition (50:50 PBAT/PLA), shown in Figure 7.27, the
micrograph displayed non-unifome shaped and dispersed droplets, much
different from the co-contiuous morphology of the uncatalyzed sample. It is
evident that both of these blends were stabilized by the catalyst addition.

Figure 7.28 shows the 30:70 PBAT/PLA blend in which PLA makes up
the continuous phase wherein PBAT droplets are dispersed. This composition
appeared to be the least affected by catalyst addition as evidenced by the both
morphology and the mechanical properities.

The observed differences. in morphology between the reactive and the
non-reactive blend compositions can be attributed to the differences in melt
viscosities of the polymer phases and to the extent of reaction between PLA and
PBAT at different blend compositions.” The transesterifcation that is expected to
have occurred during melt blending contributed to the formation of copolymers at
the interface, which, in effect, aided in suppressing coalescence, promoted ﬁnely
dispersed phases, and substantially narrowed the size distribution of the

dispersed phase.

222

 

Figure 7.26: TEM micrograph of 70:30 PBAT/PLA blend prepared with 0.05%
SnOct2

223

 

Figure 7.27: TEM micrograph of 50:50 PBAT/PLA blend prepared with 0.05%
SnOct2

224

 

Figure 7.28: TEM micrograph of 30:70 PBAT/PLA blend prepared with 0.05%
SnOct2,

The blends that showed changes in morphology with catalyst addition are
those that showed improved tensile properties from the physical blends. The
enhanced interaction of PLA and PBAT promoted better adhesion between the
phases, thus, improved stress transfer across the interface. To further examine
the relationship between the properties and their governing morphologies, a
closer look was taken at the 70:30 PBAT/PLA blend composition. The effect of
the catalyst concentration on the blend morphology development was observed.
Figures 7.29 through 7.32 depict the morphology of the blends with varying

concentrations of catalyst. Figure 7.29 shows another view of the physical blend

225

displaying the aforementioned subinclusions. From Figure 7.30 it can be seen
that the addition of a small amount of catalyst (0.05%) was sufﬁcient to break up
the droplets of PLA and hamper particle-particle coalescence while dispersing
them throughout the PBAT matrix.

An increase of catalyst (to 0.1%) resulted in a morphology with even
"smaller, more uniform particles that are better dispersed throughout the matrix
(Figure 7.31). Evidence of the relationship between the size of the dispersed
phase and the dynamic mechanical properties can be seen here. As shown by
the DMA curves (Figures 7.17 and 7.18) in the previous section, it was this
composition (containing 0.1% catalyst) that had the lowest loss and storage
moduli. The mechanical loss peaks decreased in amplitude with decreasing
phase size because of the increasing mechanical isolation of the smaller
phases.°3 The smaller phase size can also be correlated to the drastic decrease
that was seen in the storage modulus. The considerably small phase size of the
dispersed PLA in this blend (with 0.1% catalyst) allowed it to only minimally

contribute to the stiffening effect of the polymer as shown in Figure 7.18.

226

 

Figure 7.29: TEM micrograph of 70:30 PBAT/PLA physical blend.

227

 

Figure 7.30: TEM micrograph of 70:30 PBAT/PLA blend prepared with 0.05%
ShOCtz

228

 

 

Figure 7.31: TEM micrograph of 70:30 PBAT/PLA blend prepared with 0.1%
SnOct2

Surprisingly, additional catalyst did not further reduce the particle size. In
fact, upon the addition of 0.5% catalyst there was a stronger tendency for phase
coarsening. Not only did the PLA phases become larger, they also became much
less discrete, making the interface between domains less deﬁned. This system is
shown in Figure 7.32. Higher magniﬁcations of this sample permitted the phases
to be observed more clearly, as shown in Figure 7.33. This micrograph shows a
partially co-continuous morphology along with some phase inversion. This

composition also had the lowest tensile properties of any of the blends.

229

 

One reason for the phase coarsening and mechanical property decrease
could be the decrease in molecular weight with high catalyst concentration.
Fortelny et al.64 proposed a model that takes into account the classical theory of
phase break-up and the coalescence effect. According to his hypothesis a drop
in viscosity or a molecular weight decrease lowers the dispersive forces, making
phase coalescence more favorable.65 In other words the molecular weight of the
polymers controls the surface area of the interfacial region where
compatibilization takes place, thus, affecting the ultimate phase morphology of
the system. Another possible explanation Is the possibility of reaching the critical
micelle concentration, the concentration at which micelles are formed. If this
occurs the copolymer formed at the interface may migrate from this region
leading to an increase in domain size.°° Nevertheless, the morphology at this
composition suggests that a critical concentration of copolymer or compatibilizer
is required to saturate the interface of a binary blend.

Another noteworthy observation in the sample prepared with 0.5% catalyst
was the appearance of small black particles which are thought to be residual tin
particles from the added stannous octoate. These particles, which are in the
range of the size necessary to nucleate a polymer, can provide a qualitative
explanation for the increased crystallization of this blend, as observed by the
DSC data.

It is apparent that the high activity of the catalyst, in excess amounts,
could yield adverse results. The best morphology was seen in the sample with

0.1% SnOct2, however, the highest strength was seen in the sample with 0.05%

230

 

SnOct2. While it is possible that more interaction could have occurred with higher
catalyst, degradation also took place, which deteriorated the mechanical
properties. Overall these observations lend further support to the fact that these

blends were compatibilized during reactive melt blending as a result of SnOct2

addition.

 

Figure 7.32: TEM micrograph of 70:30 PBAT/PLA blend prepared with 0.5%
SI’IOCtz

231

 

Figure 7.33: TEM micrograph of 70:30 PBAT/PLA blend prepared with 0.5%
SnOct2 taken at 40,000X magniﬁcation

CONCLUSIONS

The in situ transesteriﬁcation of PBAT/PLA blends was achieved via
reactive compatibilization and the compatibilized products were successfully
blown into ﬁlm. Thermal, mechanical, and morphological properties of blends
were investigated and confirmed the use of SnOct2 to be effective for PBAT/PLA
compatibilization. In the PBAT-rich compositions a low but optimum
concentration of SnOct2 gave the highest mechanical properties with an
elongation to break of up to 350% and tensile strength improvements of nearly

190% over the physical blend. Thermal data revealed distinct transitions,

232

indicating immiscibility across the entire composition range. However, shifts in
thermal transitions suggested some interaction at the molecular level for the
catalyzed samples and enhanced nucleation at high catalyst concentrations.
Molecular weight data showed degradation of the blends with the catalyst
addition suggesting the competition between ester-exchange reactions and
degradation. Morphological observation conﬁrmed interaction between polymer

phases by improved dispersion, distribution, and reduction in domain size.

233

[$5. Ifmw

REFERENCES

1. John, J., Mani, R., Bhattacharya, M., J PolymSci Part A: PolymChem, 2002,
2003.

2. Sheth, M., Ananda Kumar, R., Dave, V., Gross, R. A., McCarthy, 8. P., 1997,
66, 8, 1495.

3. Nijenhuis, A. J., Colstee, E., Grijpma, D. W., Pennings, A. J., Polymer, 1996,
37, 26, 5849.

4. Gajria, A. M., Dave, V., Gross, R.A., McCarthy, S. P., Polymer, 1996, 37, 3,
437.

5. Park, J. W., Im, S. S., Polymer, 2003,44, 4341.

6. Eguiburu, J. L., lruin, J. J., Femandez—Berridi, M. J., San Roman, J.,
Polymer, 1998, 39, 26, 6891.

7. Jang, W.J., Shin, B. Y., Lee, T.J., Narayan, R., J. Ind. Eng. Chem., 1007, 13,
3, 457.

8. Yang, X., Yuan, M., Li, W., Zhang, G., J. Appl. Polym. Sci, 2004, 94, 1670.
9. Lee, S., Lee, J. W., Korea-Aus. Rheo. J., 2005, 17, 2, 71.
10. Ferreira, B.M.P., Zavaglia, C.A.C., Duek, E.A.R., Mater. Res., 2001, 4, 1, 34.

11.Iannace, S., Ambrosio, L., Huang, S.J., Nicolais, L., J. Appl. Polym. Sci.,
1994, 54, 1525.

12.Jiang, L., Qian, J., Wolcott, M., Zhang, J., ACS, PMSE Preprints, 2005, 93,
937.

13. Jiang, L., Wolcott, M. P., Zhang, J., Biomacromolecules, 2006, 7, 199.

14. Miksic, B., Vela, Pasko, Berg, R., Boyle, R., US. Patent 6984426, 2006.

15. Meredith, J. C., Amis, E. J., Macromol. Chem. Phys, 2000, 201, 733.

16. Broz, M.E., VanderHart, D.L.,Washbum, N.R., Biomater. 2003, 24, 23, 4181.

17.Takayama, T., Todo, M., Tsuiji, H., Arakawa, K., Trans. Jpn. Soc. Mech.
Eng., 2006, 72, 714.

18.Semba, T., Kitagawa, K., Ishiaku, U. 8., Hamade, H., j. Appl. Polym. Sci,
2006, 101, 1816.

234

19.Wang, L., Ma, W., Gross, R. A., McCarthy, S. P., Polym. Deg. Stab., 1998,
59, 161.

20.Tsuji, H., lkada, Y., J. Appl. Sci., 1995, 58, 1793.

21.Hiljanen-Vainio, M., Varpomaa, P., Seppala, J., Torrnala, P., Macromol.
Chem. Phys. 1996, 197, 1503.

22.Dell’Erba, R., Groeninckx, G., Maglio, G., Malinconico, M., Migliozzi, A.,
Polymer, 2001, 42, 7831.

23.Ege, 8., Organic Chemistry: Structure and Reactivity; Houghton Mifﬂin,
1999.

24. Paul, D. R., Newman, 8., Eds. Polymer Blends; Academic Press: San Diego,
1978,

25.Lipatov, Y.S., Nesterov, A.E., Thermodynamics of Polymer Blends; CRC
Press, 1998.

26.Walsh, D.J., Higgins, J.S., Maconnachi, A., Polymer Blends and Mixtures;
Kluwer Academic Publishers, 1985.

27. Freed, K., Phase Behavior of Polymer Blends; Birkhauser, 2005.

28.Argyropoulos, D. 0., MS. Thesis, Michigan State University, East Lansing,
MI, 1992.

29.0monov, T. S., Harrats, C., Groeninckx, G., Polymer, 2005, 46, 26, 12322.
30. White, J.L., Min, K., Macromol. Chem., Macromol. Symp., 1988, 16, 19.

31.Paul, D.R., Newman, 8., Eds., Polymer Blends; Academic Press: New York,
1978.

32.Jordhamo, Manson, Sperling, Polym. Eng. Sci., 1986, 26, 8, 517.
33. Kienzle, S. Y., Technomic Publishing, 1988, 1-6.

34.Baker, W., Scott, C., Hu, G.H., Reactive Polymer Blending; Hanser:
Connecticut, 2001.

35.Walsh, D.J., Higgins, J.S., Maconnachi, A., Polymer Blends and Mixtures;
Martinus Nijhoff: Kluwer Publishing, 1985.

36.Ajji, A., Utracki, L. A., Polym. Eng. Sci., 1996, 36, 12, 1574.

235

 

37.Zhu, S.-H., Chan, C.-M., Wong, S.C., Mai, Y.-W.,Polym. Eng. Sci., 2004, 39,
10, 1998.

38. Wang, D., Xie, X., Polymer, 2006, 47, 23, 7859.

39.Xie, W.B., Tam, K.C., Yue, C.Y., Lam, Y.C., Li, L., Hu, L., Polym. Int., 52, 5,
733.

40.Folkes, M. J., Hope, P. 8., Eds. Polymer Blends and Alloys; Chapman &
Hall: Cambridge, 1993.

41.Utracki, L. A. Can. J. Chem. Eng., 2002, 80, 1008.

42.Turi, E.A., ed., Thermal Characterization of Polymeric Materials; Academic
Press: CA, 1981.

43.ASTM Standard D 882.
44.ASTM Standard D 1938-02a.

45. Nielsen, L., Landel, R, Mechanical Properties of Polymers and Composites,
CRC Press: New York, 1994.

46.Balakrishnan, S., Ph.D. Dissertation, Michigan State University, East
Lansing, MI, 2004.

47.Solc, K., Ed. Polymer Compatibility and incompatibility: Principles and
Practices; Harwood Academic: Chur, 1982.

48.Gruber, P. R., et al., US Patent 6143863, Nov. 2000.

49.Jiang, L., Wolcott, M. F., Zhang, J., Biomacromolecules, 2006, 7, 199.
50. Bigg, D.M., SPE ANTEC Tech Papers, 1996, 2028.

51. Rim, P.B., Runt, J.P. Macromolecules, 1983, 14, 420.

52.8ivaraman, P., Manoj, N.R., Barman, S., Chandrasekhar, L., Mishra, V.S.,
Samui, A.B., Chakraborty, B.C., Polym. Testing, 2004, 23, 645—649.

53. Reed, K.C., Polym. Comp., 1980, 1, 1, 49.
54.Thomason, J.L,, Polym. Comp., 1990, 11, 2, 105.

55.Lee, S.-G., Lee, J.H., Choi, K.-Y., Rhee, J.M., Polym. Bull., 1998, 40, 765-
771.

236

m

56. Heckman, J., Flegler, S., Klomparens, K., Scanning and Transmission
Electron Microscopy: An Introduction; Oxford University Press: New York,
1995.

57.Sundararaj, U., Macosko, C. W., Shih, C. K., Polym. Eng. Sci., 1996, 36,
1796.

58.Ratnagiri, R., Scott, C. E., Polym. Eng. Sci., 1998, 38, 1751.
59.Scott, C.E., Joung, S. K., Polym. Eng. Sci., 1996, 36, 1666.

60.Harrats, C., Thomas, S., Groeninckx, G., Micro- and Nanostructured
Multiphase Polymer Blend Systems: Phase Morphology; CRC Press: New
York, 2005.

61.Favis, B., Chalifoux, J., Polymer, 1988, 29, 1761.
62.Astruc, M., Navard, P., J. Rheo., 2000, 44, 693.
63. Wetton, R. E., Moore, J. D., Ingram, P., Polymer, 1973, 14, 161.

64.Jerome, R., Pagnoulle, C., Key Role of Structural Features of
Compatibilizing Polymer Additives in Reactive Blending In Reaction Polymer
Blending, Eds. Baker, W., Scott, C., Hu, G.-H.; Hansel“. Munich, 2001, 83.

65.Fortelny, |., Cema, Z., Binho, J., Kovar, J., J. Apl. Polym. Sci., 1993, 48,
1731.

66.George, S., Neelakantan, R., Varughese, K.T., Thomas, 8., J Polym. Sci. B:
Polym. Phys, 1997, 35, 2309.

237

Chapter 8

 

CONCLUSIONS AND RECOMMENDATIONS

 

CONCLUSIONS

In this contribution it has been shown that PLA-based systems are
compounds of rather great potential for applications requiring biobased content,
biodegradation, and sufﬁcient properties. The goal of this work, as illustrated in
Chapter 1, was to develop new PLA-based materials by reactive modiﬁcation in
an effort to overcome some of the inherent drawbacks of PLA. Chapters 4
through 7 describe in detail the work done towards meeting this goal. Detailed
conclusions for each project are summarized in its respective chapter, however a
general summary is provided below.

The study reported in Chapter 4 of this dissertation showed that maleic
anhydride-grafted PLA can be initiated by the use of two structurally different

peroxides—L101 and T301. It was reported that T301 yielded products with

238

higher molecular weight when compared to L101. It was also discussed that
there exists an optimum monomer-to-initiator ratio that should be employed in
order to obtain the correct balance between grafting and molecular weight.
Nevertheless, the maleated copolymer was shown to be effective when used as
an interfacial modiﬁer in PLA-talc composites as presented in Chapter 5. In that
chapter, the inﬂuence of MAPLA on the mechanical and thermal behavior of the
composites was investigated. It was found that MAPLA, when added in an
optimum concentration, improved the tensile strength and crystallization of the
composite. Furthermore, the thermomeChanical analysis together with
microscopic observation conﬁrmed the compatibilization effect of MAPLA in PLA-
talc composites. In Chapter 6 of this dissertation, it was demonstrated that PLA
can be functionalized with Vinyltrimethoxysilane and subsequently crosslinked in
the presence of moisture. Thermal and mechanical characterization revealed that
the crosslinking increased the thermal stability while maintaining the mechanical
strength of PLA. In the ﬁnal project discussed in Chapter 7 blown ﬁlms
comprising of reactively compatibilized PLA-PBAT blends were produced. The
ﬁlms exhibited enhanced interfacial interaction when compared to the physical
blends, shown by mechanical property improvements as well as enhanced
dispersion and reduced domain size of the dispersed phase. Moreover, the

results conﬁrmed the compatibilization effect of the reaction.

239

 

 

RECOMMENDATIONS FOR FUTURE WORK

Maleated Polylactic Acid (MAPLA) as an Interfacial Modiﬁer in Polylactic Acid-
Talc Composites

Mechanical property determination of PLA/MAPLA/Talc composites
containing varying amounts of talc

Compounding PLA/talc composites using a MAPLA masterbatch
formulation to minimize thermal degradation of MAPLA

Reactive Modiﬁcation of Polylactic Acid by Silane Grafting and Crosslinking
Reactions

Viscoelastic study of the crosslinked PLA to determine the effects of
crosslinking on the rheological properties

Determination of grafting yield by FT IR peak deconvolution and
supplemented by elemental analysis techniques

NMR study of silane-grafted polymer

Incorporation of bioﬁbers into silane-crosslinked PLA to impart more
biobased content and the investigation of the thermal and mechanical
properties of the composite for applications requiring thermal stability

Use of crosslinked PLA as a self-reinforcing agent into linear PLA

Reactive Compatibilization of Polylactic Acid and Polybutylene Adipate co-
Terephthlate Blends

Attainment a suitable solvent to selectively extract one component of the
polymer blend to determine relative degree of reaction and further
characterization of copolymer

Deconvolution of GPC peaks to accurately determine the molecular weight
of each component

Water vapor and oxygen permeability study of compatibilized PLA/PBAT
blown ﬁlms

Use of organic peroxides to compatibilize PBAT and PLA by crosslinking
reactions

240

LIBRARIES

II

84

03063 2

Y
n
S
R
E
W
N
U
E
T
A

11111

MICHIGAN