a . A): .
ii .15. a
z... u ..:. . 1

‘5’! l.. 1x
3, 1.843.
2.

£9595?"

,9;ng tum-‘qumy

O

m

«an. "‘39” Psghfflfl”

an...
._ mmemuwmm
:mmnfivdmd any.

I I O...

V... .x
Eui 3

M ‘
£3

9
2’

g. 5

5,

31
$3.," ..

I.
1:1 ‘35.
4.324;.
.1

.fl 3.. Ami.”

Lu. .. z i.

‘15
hung . ‘

1: n1.
at}. 1.5

.u... ; 11x1... ‘ inc!
2.... r i .

 

 

iii-"5'“ LIBRARY
Michigan State
University

 

 

 

This is to certify that the
dissertation entitled

NUCLEATING EFFECT OF EXFOLIATED GRAPHITE
NANOPLATELETS ON POLY(HYDROXYBUTYRATE) AND
POLY(LACT|C ACID) AND THEIR NANOCOMPOSITES
PROPERTIES

presented by

DANA GABRIELA MILOAGA

has been accepted towards fulfillment
of the requirements for the

Doctor of

Philosophy degree In Chemical Engineering

 

 

Major Professor’s Signatufé
0% /5« 0:

Date

 

MSU is an affirmative—action, equal-opportunity employer

—~-o-t-¢-a-0--------o-o-o-o-o---I---u-o-I-u-

.c-----t’.-u-u-u-—-.—--w-o--n-n--.-o-c-n-c-o-o-o-c---I_I-.-v

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.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K'IProilAcc8PrelelRC/DateDue.indd

NUCLEATING EFFECT OF EXFOLIATED GRAPHITE NANOPLATELETS on
POLY(HYDROXYBUTYRATE) AND POLY(LACTIC ACID) AND THEIR
NANOCOMPOSITES PROPERTIES

By

Dana Gabriela Miloagé

A DISERTATION

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

DOCTOR OF PHILOSOPHY

Chemical Engineering

2008

ABSTRACT

NUCLEATING EFFECT OF EXFOLIATED GRAPHITE NANOPLATELETS ON
POLY(HYDROXYBUTYRATE) AND POLY(LACTIC ACID) AND THEIR
NANOCOMPOSITES PROPERTIES

By
Dana Gabriela Miloagé

Multifunctional polymer nanocomposites are attractive due to the multitude of
properties that can be imparted to polymer matrices by the addition of small
amounts of nanoparticles. Exfoliated graphite nanoplatelets (xGnP) represent a
new nanosize reinforcing material that can lead to multifunctional polymer
nanocomposites, and are a significantly less expensive alternative to other
carbon nanomaterials (e.g. carbon nanotubes). The benefits of xGnP have been
shown in nanocomposites made with several petroleum-derived thermoplastic
polymers. Nanocomposites from biobased polymers and xGnP have not yet
been investigated and are the object of this study.

This research is directed at understanding the interactions between xGnP and
two polyesters derived from renewable resources, poly(3-hydroxybutyrate)
(PHB), and polylactic acid (PLA). PHB and PLA are both semicrystalline
polymers. Their properties largely depend on the degree of crystallinity,
crystallization rate, and on the size and number of crystallites formed. For faster
crystallization, important especially for dynamic processing conditions, both

polymers often require the addition of nucleation agents. xGnP is shown to act as

a nucleating agent, and to impart higher strength, modulus, thermal and electrical
conductivity to these polymer matrices as well. Specifically, the research
conducted covers two main aspects of the nucleating effect of xGnP on PHB and
PLA: the rate of overall crystallization from the melt as a function of the amount
of xGnP, and the influence of xGnP on the morphology and properties of the
nanocomposites. A kinetic study of the PHB/xGnP crystallization, employing a
modified form of the Avrami equation, revealed that the presence of xGnP
increased the PHB crystallization temperature, as well as the crystallization rates.
Optical microscopy and atomic force microscopy investigations revealed the
reduction in the size of PHB spherulites. Nanocomposite properties were studied
for PLA/xGnP systems, with a special focus on the relation between the
nucleating effect of xGnP and improvement in modulus, and thermal conductivity.
The results are particularly important since they may be applied to

nanocomposites utilizing similar semicrystalline polymer matrices.

Copyright by
DANA GABRIELA MILOAGA
2008

To my parents,

Leti and Stefan Miloagé

ACKNOWLEDGEMENTS

It is with great pleasure that my first acknowledgement goes to Professor
Lawrence T. Drzal. I would like to thank him for making possible the special mix
of research freedom, personalized guidance, encouragement, support, work and
enjoyment which was my graduate school experience. Many thanks also to my
Ph.D. committee members, professors Ramani Narayan, Gregory Baker, and
Andree Lee, and to the Department of Chemical Engineering and Materials
Science.

Special thanks Dr. Manju Misra, Dr. Hazel-Ann Hosein, Dr. Per Askeland,
Brian Rook, Robert Jurek, Kelby Thayer, Ed Drown, and Mike Rich in the
Composite Materials and Structures Center for a safe, professional, and
instructive work environment, and for their friendship along the years.

For encouragement, bouncing ideas, solving problems, coffee breaks, fun,
books and movies, I would like to thank everyone who shared the office in 1105
as a graduate student, my friends in the Narayan group, Dr. Wanjun Liu, and
Toshiya Kamae.

A big thank you goes to Nathan, who helped me put all this work in
perspective and keep my feet on the ground.

This work is dedicated to my parents — symbols of love and sacrifice.

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................... viii
LIST OF FIGURES ............................................................................................... x
CHAPTER 1
INTRODUCTION AND BACKGROUND ............................................................... 1
Current trends in polymer nanocomposites ................................................ 1
Market trends and applications ........................................................ 1
Design challenges and recent research focus ................................. 3
Description of the present study ................................................................. 7
Context and motivation .................................................................... 7
Objective .......................................................................................... 8
Approach ......................................................................................... 9
Background on nucleation of semicrystalline polymers ............................ 10
The importance of nucleation for semicrystalline polymers ........... 10
The heterogeneous nucleation of crystalline polymers .................. 11
Nucleating agents for semicrystalline polymers ............................. 13
Factors that reduce crystallinity in polymers .................................. 21
Nanoparticles nucleating semicrystalline polymers ....................... 23
Nanocomposites based on biodegradable polymers ................................ 26
Poly(hydroxybutyrate) and polylactic acid polymers ...................... 27
The importance of nucleation for PHB and PLA ............................ 30
Exfoliated graphite nanoplatelets as nucleating agents ............................ 42
References ............................................................................................... 45
CHAPTER 2
NUCLEATION OF POLY(HYDROXYBUTYRATE) BY EXFOLIATED
GRAPHITE NANOPLATELETS .......................................................................... 57
Introduction .............................................................................................. 57
Experimental details ................................................................................. 59
Results and discussion ............................................................................. 62
Conclusions .............................................................................................. 81
References ............................................................................................... 82
CHAPTER 3
MODIFICATION OF CRYSTALLINITY AND xGnP/PHB INTERACTION
WITH IONIC LIQUIDS IN PHB/xGnP NANOCOMPOSITES ............................... 85
Introduction .............................................................................................. 85
Effects of two imidazolium phosphates on PHB ....................................... 87
Experimental details ...................................................................... 89
Results and discussion .................................................................. 92
Conclusions ................................................................................... 98
PHB/1-butyl-4-methy|imidazolium acetate/xGnP nanocomposites ........... 98

vii

Experimental details .................................................................... 100

Conclusions ................................................................................. 1 1 1
References ............................................................................................. 1 12
CHAPTER 4
POLYLACTIC ACID/EXFOLIATED GRAPHITE NANOPLATELET
NANOCOMPOSITES ........................................................................................ 1 14
Introduction ............................................................................................ 1 14
Experimental details ............................................................................... 1 15
Results and discussion ........................................................................... 121
Conclusions ............................................................................................ 133
References ............................................................................................. 134
CHAPTER 5
THERMAL CONDUCTIVITY OF POLYLACTIC ACID/ EXFOLIATED
GRAPHITE NANOPLATELET NANOCOMPOSITES ....................................... 135
Introduction ............................................................................................ 1 35
Experimental details ............................................................................... 140
Results and discussion ........................................................................... 144
Conclusions ............................................................................................ 151
References ............................................................................................. 1 53
CHAPTER 6
SCANNING PROBE THERMAL ANALYSIS OF POLYLACTIC ACID/
EXFOLIATED GRAPHITE NANOPLATELET NANOCOMPOSITES ................ 154
Calibration standards and testing conditions for the study of
PHB/xGnP and PLA/xGnP nanocomposites .......................................... 155
Introduction .................................................................................. 1 55
Experimental details .................................................................... 160
Results and discussion ................................................................ 162
Conclusions ................................................................................. 169
Experimental investigation of polylactic PLA/xGnP nanocomposites ..... 170
Introduction .................................................................................. 1 70
Experimental details .................................................................... 1 70
Results and discussion ................................................................ 174
Conclusions ................................................................................. 188
References ............................................................................................. 1 89
CHAPTER 7
CONCLUSIONS ................................................................................................ 193

viii

Table 1.1
Table 2.1
Table 3.1
Table 3.2

Table 3.1
Table 3.2

Table 3.3

Table 3.4

Table 3.5

Table 3.6

Table 5.1

Table 5.2

Table 6.1

Table 6.2

Table 6.3

Table 6.4

LIST OF TABLES

Classification of biodegradable polymers ...................................... 27
Summary of modified Avrami analysis results ............................... 65
Common cations and anions in ionic liquids .................................. 87
Equilibrium melting temperatures of PHB samples containing

various plasticizing agents ............................................................. 96
Common cations and anions in ionic liquids .................................. 87

Equilibrium melting temperatures of PHB samples containing
various plasticizing agents ............................................................. 96

Processing conditions and forces recorded by the DSM
extruder for PHB, and PHB containing lL3 .................................. 104

Dilution scheme and dilute solution viscometry results for
PHB in chloroform ....................................................................... 106

Dilution scheme and dilute solution viscometry results for
PHB containing 1 wt % IL3 in chloroform ..................................... 106

Atomic concentration for exfoliated graphite nanoplatelets
surface treated with IL3 (XPS results) ......................................... 109

Room temperature thermal conductivity of materials of
various crystalline structures ....................................................... 138

Properties of various nanosize fillers used for improving
thermal and electrical conductivities of polymers ......................... 139

Melting temperatures of PCL, PE, PET and Pro-fax 6301
obtained from DSC and nano-TA experiments ............................ 165

Melting temperatures recorded by nano-TA using scan rates
from 60 to 600 °C/min .................................................................. 169

Summary of DSC analysis of melting and nonisotherrnal
crystallization from the melt of PLA/xGnP-1 composites ............. 180

Summary of the DSC analysis of cold crystallization of
PLA L209S and PLA L21OS for different heating rates ................ 184

Figure 1.1

Figure 1.2
Figure 1.3

Figure 1.4

Figure 1.5
Figure 1.6
Figure 1.7
Figure 1.8
Figure 1.9

Figure 1.10

Figure 1.11

Figure 1.12

Figure 1.13

Figure 2.1

LIST OF FIGURES

Schematic of the contribution of interfacial layer to the
properties of PNC ............................................................................ 6

Schematic representation of primary nucleation ............................ 12

Effect of nucleating agent type and content on the
crystallization temperature of PP matrix. (D) CaCO3; (El) talc ....... 17

Dependence of the crystallization temperature on the cooling
rate for PP, and PP containing CaCO3 (CC), talc (TC),
p-tert-dibutyl benzoic acid monohydroxy aluminium (BA),

and p—dimethyI-benzylidene sorbitol (GA) ...................................... 17
Behavior of nucleating and clarifying agents ................................. 19
Effect of the size and amount of talc particles on PP ..................... 20
Chemical structure of PHAs ........................................................... 28
Synthesis of polylactides ............................................................... 29

DSC curves of quenched PLLA/OMLS blends containing
different amounts of OMLS ............................................................ 38

DSC traces of neat PLA and PLA/CN. (b) Polarized optical
micrographs of crystallized (at 110 °C) samples of neat PLA
and PLAICN ................................................................................... 40

Dependence of the overall crystallization rate of PLA, and
PLA/OMLS nanocomposites on the crystallization

temperature ................................................................................... 41
Boron nitn'de crystal structures ...................................................... 42
Schematic of the xGnP .................................................................. 43

Nonisotherrnal crystallization curves obtained for (a) pure
PHB, (b) PHB/1% xGnP-1 and (c) PHB/3% xGnP-1 at six
cooling rates: 5°C/min; 8°C/min; 10°C/min; 14°C/min;
17°C/min; and 20°C/min respectively. (The arrows indicate
increasing cooling rates.) ............................................................... 63

Figure 2.2 Graphs showing (a) the temperature of crystallization peaks
and (b) the degrees of supercooling versus cooling rate for
pure PHB (I), PHB/0.01% xGnP-1 (D), PHB/1% xGnP-1 (o),
and PHB/3% xGnP-1 (A). ............................................................. 64

Figure 2.3 Development of relative crystallinity with temperature for (a)
pure PHB, (b) PHB/1% xGnP-1, (c) PHB/3% xGnP-1 at six
cooling rates: 5°C/min; 8°C/min; 10°Clmin; 14°C/min;
17°C/min; and 20°C/min, respectively. (The arrows indicate
increasing cooling rates.) ............................................................... 68

Figure 2.4 Development of relative crystallinity with time for (a) pure
PHB, (b) PHB/1% xGnP-1, (c) PHB/3% xGnP-1 at six
cooling rates: 5°C/min; 8°C/min; 10°Clmin; 14°C/min;
17°C/min; and 20°C/min respectively (The arrows indicate
increasing cooling rates.) ............................................................... 69

Figure 2.5 Graphs showing the results of modified Avrami analysis for
(a) PHB, (b) PHB/1% xGnP-1, and (c) PHB/3% xGnP-1 at
six cooling rates ............................................................................. 70

Figure 2.6 Nonisotherrnal crystallization rate parameters for PHB/xGn-1
systems, determined according to modified Avrami analysis ......... 71

Figure 2.7 Graphs showing (a) the half-time crystallization times and (b)
the maximum crystallization times for pure PHB (I),
PHB/0.01% xGnP-1 (El), PHB/1% xGnP-1 (~), and PHB/3%
xGnP-1 (A), determined according to modified Avrami
analysis .......................................................................................... 72

Figure 2.8 Kissinger plots used for determining the activation energy for
nonisothermal crystallization of pure PHB (I), PHB/0.01%
xGnP-1 (El), PHB/1% xGnP-1 (o), and PHB/3% xGnP-1 (A) ......... 73

Figure 2.9 Activation energies for the nonisothermal crystallization from
melt of PHB/xGnP-1 systems ........................................................ 74

Figure 2.10 Optical micrographs showing spherulites of (a) pure PHB
and (b) PHB/0.01 wt % xGnP-1 formed during nonisothermal
crystallization from the melt. . ........................................................ 75

Figure 2.11 AFM tapping mode height images (top view) showing
spherulites of (a) pure PHB and PHB containing, (b) 0.01 wt
% xGnP-1, (c) 1 wt % xGnP-1 and (d) 3 wt % xGnP-1 .................. 76

xi

Figure 2.12

Figure 2.13

Figure 2.14

Figure 3.1
Figure 3.2

Figure 3.3

Figure 3.4

Figure 3.5

Figure 3.6

Figure 3.7

Figure 3.8

Figure 3.9
Figure 3.10

Figure 3.11

Figure 3.12

Figure 3.13

AFM tapping mode height images (top view) and
accompanying cross-sections showing the Iamellar
thicknesses of spherulites of (a) pure PHB and (b) PHB/1 %
wt xGnP. ........................................................................................ 77

AFM height image (top view) of PHB/3 % xGnP-1 showing
the apparent disruption of the Iamellar structure of the

polymer .......................................................................................... 78
Summary of WAXD analysis of PHB/xGnP-1

nanocomposites ............................................................................ 80
Structures of a) lL1, and b) II2 ....................................................... 90

Flexural strength of PHB samples containing plasticizing
agents ............................................................................................ 93

Flexural modulus of PHB samples containing plasticizing
agents ............................................................................................ 93

Notched Izod impact strength of PHB samples containing
plasticizing agents ......................................................................... 94

Glass transition temperatures of PHB samples containing
various plasticizing agents ............................................................. 94

DSC data collected for the determination of equilibrium
melting temperatures of PHB containing various potential
plasticizers ..................................................................................... 96

Extrapolated DSC data for PHB containing lL1, IL2, and TEC ...... 97
DSC traces for PHB and PHB containing IL3 .............................. 105
Determination of the intrinsic viscosity of processed PHB ........... 107

Determination of the intrinsic viscosity of processed PHB
containing 1 wt % IL3 .................................................................. 108

Effect of surface treatment of xGnP-1 with IL3 on the flexural
strength of PHB/xGnP-1 nanocomposites ................................... 110

Effect of surface treatment of xGnP-1 with IL3 on the
modulus of PHB/xGnP-1 nanocomposites ................................... 110

Effects of surface treatment of xGnP-1 with IL3 on the
notched Izod impact strength of PHB/xGnP-1
nanocomposites .......................................................................... 1 10

xii

Figure 4.1

Figure 4.2

Figure 4.3

Figure 4.4

Figure 4.5

Figure 4.6

Figure 4.7

Figure 4.8

Figure 4.9

Figure 4.10

Figure 4.11

Figure 4.12

Figure 4.13

Optical micrographs showing PLA/xGnP-1 composites
crystallized from the melt under isothermal conditions. The
amount (wt %) of xGnP-1 was a) 0 %; b) 0.01 °/o c) 1 % and
d) 10% ......................................................................................... 122

Degree of crystallinity of PLA/xGnP-1 composites prepared
by extrusion and injection molding ............................................... 123

Flexural moduli of PLA 30010 and PLA 3001D [xGnP-1
nanocomposites prepared by extrusion and injection molding ....125

Flexural strength of PLA 3001D and PLA 3001D /xGnP-1
nanocomposites prepared by extrusion and injection molding ....125

Bulk electrical conductivity of injection molded PLA 3001D
/xGnP—1 nanocomposites ............................................................ 126

Moduli of PLA L21 OS and PLA L21OS/xGnP-1 composites
prepared by pre-mixing and compression molding, as
determined by DMA ..................................................................... 127

Moduli of PLA L21OS and PLA L21OS/xGnP-1 composites,
as determined from nanoindentation experiments ....................... 128

Surface hardness of PLA L21OS and PLA L21OS/xGnP-1
composites, as determined from nanoindentation
experiments ................................................................................. 129

DSC traces for the nonisothermal crystallization from the
melt of PLA L21OS, and PLA L21OS containing xGnP-1 ............. 130

Areas of peaks corresponding to nonisothermal
crystallization from the melt of PLA L21OS, and PLA
L21OS/xGnP-1 composites .......................................................... 131

Full-width-at-half-maximum values of the crystallization
peaks corresponding to PLA L21OS, and PLA L21OS/xGnP-1
composites .................................................................................. 1 31

Degree of crystallinity of PLA/xGnP-1 composites prepared
by pre-mixing and compression molding ..................................... 132

Storage moduli of compression molded PLA L21OS and PLA
L21OS/xGnP-1 composites prepared using two mixing
approaches .................................................................................. 134

xiii

Figure 5.1

Figure 5.2

Figure 5.3

Figure 5.4

Figure 5.5

Figure 5.6
Figure 5.7

Figure 6.1

Figure 6.2

Figure 6.3
Figure 6.4

Figure 6.5

Figure 6.6

Figure 6.7

Localization of carbon black particles (1 % wt) in the PP
phase of a PP/ PS blend (a), and at the interface between
PP and PS in a similar blend (b) .................................................. 141

Room temperature thermal conductivity of extrusion (melt)
mixed injection molded PLA 3001 D/xGnP-1 composites ............. 145

Room temperature thermal conductivity of PLA L21OS/xGnP
composites prepared by solution mixing and compression
molding ........................................................................................ 146

Specific electrical conductivity of PLA L21OS/xGnP-1
composites prepared by solution mixing and compression
molding ........................................................................................ 147

Optical micrographs of a) PLA L21OS containing 10 %
xGnP-1 and b) composite sample containing 70 % PLA
L21OS, 20 % PS, and 10 % xGnP-1 ............................................ 150

Thermal conductivity of PLA/PS/xGnP-1 composites .................. 151

Storage moduli of PLA, PS, and PLAIPS/xGnP-1 composites
containing 1 % xGnP-1 ................................................................ 152

Scanning Electron Micrographs of a first generation nano-TA
probe showing a) entire cantilever; b) the 1pm tip) and c) a
temperature map of the probe (Images courtesy of Anasys
Instruments, CA) .......................................................................... 1 58

AFM images of a PLA surface before nano-TA a) height
image, and b) deflection image and c) after nano-TA,
showing the indentation formed when the sample melts
(Scans are all 3 pm2 and collected at a scan rate of 2 Hz.) ........ 159

Example of nano-TA thermal profile for polyhydroxybutyrate ...... 160
DSC curves showing the melting behavior of a) PCL; b) PE

and c) PET ................................................................................... 166
Summary of DSC results showing Tm, Tmp, Tm(lit), and
Tm(nTA) values for PCL, PE, PET and Pro-fax ........................... 168
Nano-TA results showing the effect of heating rate on the
melt temperature for a) Pro-fax 6301, b) PE and c) PET ............. 170
Storage moduli of PLA] xGnP-1 composites ................................ 177

xiv

Figure 6.8 DSC analysis of the nonisothermal crystallization from the
melt of (a) PLA L209S and (b) PLA 2108 and their
composites with xGnP 1 .............................................................. 179

Figure 6.9 Optical micrographs showing spherulites of (a) pure PLA
L21OS; (b) PLA L21OSI 0.01 % % wt xGnP-1 and (c) PLA
L21OS/ 5 % wt xGnP-1 formed during nonisothermal
crystallization from the melt. The scale bar is 20 um. .................. 182

Figure 6.10 Scanning probe themIal analysis deflection curves for a)
PLA L21OS/xGnP-1 and b) PLA L209$lenP-1 composites ....... 186

Figure 6.11 DSC therrnograms showing the cold crystallization and
melting of a) PLA L209S and b) PLA L21OS at three heating
rates ............................................................................................. 189

Figure 6.12 DSC therrnograms showing the melting of a) PLA

L209S/xGnP-1 and b) PLA L21OS/xGnP-1 composites
containing increasing amounts of xGnP-1 ................................... 190

XV

AFM
BN
CNT
DMA
DSC
HDPE
IL

|L1

|L2
IL3

LS
M-H-S
MWNT

na no-TA

OM
OMLS
PDLA
PE
PEC
PET
PHA
PHA

KEY TO SYMBOLS OR ABBREVIATIONS

atomic force microscopy

boron nitride

carbon nanotube

dynamic mechanical analysis

differential scanning calorimetry

high density polyethylene

ionic liquid

1-ethyl-3-methylimidazolium hexafluorophosphate
1-butyl-3-methylimidazolium hexafluorophosphate
1-butyl-4-methylimidazolium acetate

light scattering

Mark-Houwnik-Sakurada

multi-walled carbon nanotubes

nanoscale thermal analysis (scanning probe thermal
analysis)

optical microscopy

Organically modified layered silicates
poly(D-lactic acid)

polyethylene

poly(e-caprolactone)

poly(ethylene terephthalate)
poly(hydroxyalkanoate)
poly(3-hydroxyalkanoate)

xvi

PHB
PLA
PLLA
PMMA
PNC
POM
PP
PVC
Stdev.
SWNT
TEC
TMA
RTIL
xGnP
xGnP-1

xGnP-15

XPS

poly(3-hydroxybutyrate)
polylactic acid

poly(L-Iactic acid)

poly(methyl methacrylate)
polymer nanocomposites
polarized light optical microscopy
polypropylene

poly(vinyl chlorite)

Standard deviation

single-walled carbon nanotubes
triethyl citrate

therrnomechanical analysis
room temperature ionic liquids
exfoliated graphite nanoplatelets

exfoliated graphite nanoplatelets of 1 mm diameter

(average)

exfoliated graphite nanoplatelets of 15 mm diameter
(average)

X-Ray Photoelectron Spectroscopy

xvii

CHAPTER 1

INTRODUCTION AND BACKGROUND

Current trends in polymer nanocomposites

Polymer nanocomposites (PNC), obtained by incorporation of nanosize
particles into polymer matrices, have attracted a substantial research interest
from academic and industrial organizations during the past decade."3 Small
amounts (up to 5 % vol) of nanosize particles - spheres, rods, or platelets — were
shown to substantially modify the properties of both therrnosets and
therrnoplastics, revealing advantages particularly for applications requiring

enhanced properties and light weight.

Market. trends and applications

According to a market research study performed by The Freedonia Group,
Inc., a Cleveland-based industry research firm, by 2010 the nanocomposites
demand will grow to nearly 350 million pounds per year, and will rise to over
seven billion pounds with a value nearing $15 billion by 2020. Simultaneously,

thermoplastic polymers are estimated to account for eighty percent of the

predicted total demand for PNC in 2020.4 In the shorter term, the report
estimates that demand growth will be fastest in higher-priced resins such as
engineered plastics and thermoplastic elastomers, as much of the initial demand
for nanocomposites will be in higher-end applications. Eventually, polymer
nanocomposites based on commodity plastics, such as polypropylene (PP),
polyethylene (PE), poly(vinyl chloride) (PVC) and polyesters are expected to
dominate the market. An emerging area will be nanocomposites from biobased
polymers, which are of significant importance considering the general effort to
reduce the amount of plastic waste generated, and the recent focus on
biodegradable polymers as potential alternatives to traditionally used materials
like polyethylene terephthalate (PET) and polyoletins.5

Organically modified layered silicates (OMLS) and carbon nanotubes (CNT)
continue to dominate the market, with OMLS being extensively studied for
applications in the packaging industry. OMLS added to polymers can impart
enhanced stiffness at low addition levels, can improve barrier properties, heat
distortion temperature, and flame-retardancy of polymer matrices, producing
nanocomposites with commercial applications in a variety of diverse markets
ranging from automotive to packaging.6 In addition to improving mechanical
properties, CNT and carbon-based nanoreinforcements distinguish themselves
by their ability to impart thermal and electrical conductivity to polymers.

The automotive industry is still leading the Way in nanocomposite application
development”, with electrically conductive PNC becoming preferred materials

for fuel delivery lines, traditionally made from steel. Hyperion Catalysis

International estimates that 60 % of the cars produced in the United States in
2007 incorporated their multi-walled carbon nanotubes (MWNT) composites in
fuel lines to prevent the accumulation of static electricity.7 Electrically conductive
materials incorporating CNT were also developed for coatings used in exterior
body panel applications. Higher-end applications in the automotive industry
include batteries, fuel cells, energy storage materials, and materials for
electromagnetic and radiofrequency shielding.

The electronics and medical areas are two other sectors largely developing
PNC requiring improved electrical and thermal conductivities, for applications
ranging from thin-film capacitors in integrated circuits, solid polymer electrolytes
for batteries, nanoscale switches and sensors, and minimally invasive diagnosis
devices.

There is a strong need for PNC with enhanced thermal and electrical
conductivity, and nanosize reinforcements that can be alternatives to the still

highly-priced CNT are highly desired.

Design challenges and recent research focus

Regardless of the type of nanosize reinforcements, the challenges in PNC
differ from common challenges encountered in traditional composites.
Dimensional similarities between physical characteristics of polymers (such as
radius of gyration and Iamellar thickness) and nanosize reinforcing particles lead
to the necessity to go beyond filled systems8 and explore in detail

polymer-particle interactions responsible for the improved properties of PNC.

Such interactions, together with the use of external forces and gradients during
processing, are more likely to be responsible for the properties of PNC than
dispersion and alignment — significant factors considered in composites
containing micron-sized fillers.

The success achieved by Toyota in the 1980s with Nylon/clay

nanocompositesg‘m' ‘1

set the benchmark for the use of OMLS for improving
mechanical, thermal and physical properties of polymer matrices by controlling
the exfoliation of clay layers. Eariier research on nanocomposites incorporating
OMLS was directed predominantly at the understanding of dispersion and
exfoliation of nanoplatelets. Important research was also focused on processing
methods for PNC, from in-situ polymerization to polymer intercalation from
solution and polymer melt intercalation. It was in the late 19903 and early 2000s
when a large number of research groups directed their attention to modifications
induced by OMLS nanoplatelets to the crystalline structures of the host polymers.
Polymorphism was investigated in relation to incorporating OMLS in
Nylon12‘13‘14’17, syndiotactic polystyrene”, poly(vinyl alcohol) (PVA)19. At the
same time, with nanoscale morphological characterizations becoming more
accessible through transmission electron microscopy (TEM), atomic force
microscopy (AFM), and wide angle X-ray diffraction (WAXD), the effects of
OMLS on the shape and size of crystallites formed by various polymers was also
an important research field. Reports on the crystallization, properties, and crystal

and nanoscale morphology of PET/OMLS nanocomposites were published as

early as 1999. Concomitant work reported by Nam et al.20 investigated the

hierarchical structure of the PP/OMLS systems using wide-angle X-ray diffraction
(WAXD), small-angle X-ray scattering (SAXD), TEM, polarizing light optical
microscopy (POM) and light scattering (LS), focusing on the modifications of PP
Iamellar thickness and spherulitic structure.

By 2005, it was evident that for semicrystalline polymers nanocomposites
polymorphism, changes in crystallites shape and size, as well as modification of
the time required to achieve maximum crystallinity were general effects resulting
from the addition of nanoparticles to polymers.21

During the last five years, the physical aspects of PNC related to polymer
behavior when subjected to confinement by the presence of nanoparticles with
large surface area, and the specific interactions occurring at

22. The contribution of

polymer-nanoparticles interfaces were key issues3'
polymer-particle interfaces to the properties of the nanocomposites increases as
the number of the reinforcing particles increases. The interface and the
interaction between particles dominate the macroscopic properties.3 For
example, in case of platelets of diameter d and thickness h (Figure 1.1),
supposing that the thickness (t) of interfacial layer that surrounds the particle is
independent of the size of the particle, the relation between the particles’

contribution and the interface’s contribution to the bulk properties of the

nanocomposites is given by Equation 1.1.

232(g+h)t
Vinterface = 2 2 =Et_(1+.&) (11)
d .

vparticle ”(if h h
2

 

interfacial layer surrounding the
particle (thickness 1)

 

platelet (diameter d and thickness h.

 

 

 

 

 

Figure 1.1 Schematic of the contribution of interfacial layer to the properties of
PNC

In a recent review”, Vale and Maguire emphasized the opportunities afforded
by the ability to control nanoparticles hierarchy, and underlined the two general
approaches to this challenge: a) extemaI-in (directed patterning of nanoparticles
dispersion), and b) intemal—out (mesophase assembly of nanoparticles).

One general trend in currently designed PNC involves careful evaluation of
polymer-particle interface, and the use of chemical and physical methods to alter

the interface and thus control material properties.

Description of the present study

Context and motivation

The diversity of PNC is dramatically increasing as new nanosize
reinforcements are developed, and as the list of polymer matrices is longer
limited to PP and Nylon, as in the early years of PNC. The types of polymer
matrices expanded, including a large variety of epoxies, engineered
therrnoplastics and liquid crystalline polymers.

In the fast-developing research on PNC, during the last ten years there has
been intensive research on biobased, biodegradable polymers as matrices,
especially for composites with OMLS, in order to reduce the effects of plastic
waste on the environment. The United States generates more than 14 million
tons of plastic waste annually23' 24, only 28 % of which is recycled at a cost of
several million dollars”. According to the principles of green chemistry, there is
a need for polymeric materials whose production would not involve the use of
toxic products, and that could be degraded to natural, nontoxic products.

The transition to biobased materials will be enhanced if new and desirable
attributes can be added to their intrinsic properties. In addition to an increasing
demand for PNC with high mechanical properties, materials with improved
thermal and electrical conductivity are increasing in importance, so imparting
electrical and thermal conductivity to biobased polymer matrices is of great

interest.

The exfoliated graphite nanoplatelets (xGnP) developed in the Drzal Group at
Michigan State University represents a novel attractive carbon-based nanosize
material which when added to and dispersed within polymers at a concentration
of a few percent, adds multifunctionality as well. While significantly less
expensive than other carbon nanomaterials (e.g. CNT), xGnP were already
shown to be comparable to those when used in nanocomposites with several
petroleum-derived thermoplastic polymers, such as Nylon26 and PP.”30

Nanocomposites from biobased polymers and xGnP have not been
investigated, though they are particularly attractive for applications ranging from
structural materials to substrates in electronics, advanced energy sources, and
even biomedical applications such as tissue engineering, bone implants and
bone growth stimulation, which require biocompatible materials with enhanced

mechanical properties31 and improved conductivity.

Objective

It is the objective of this study to identify and characterize the effects of the
exfoliated graphite nanoplatelets on the crystallization of poly(hydroxybutyrate)
(PHB) and polylactic acid (PLA) polymers and the properties of their resulting

nanocomposites.

Approach

For semicrystalline polymer matrices, the nucleation by nanoparticles is one
particular interaction that affects the polymer crystallization and further the
properties of the resulted nanocomposites.

PHB and PLA are both semicrystalline polymers derived from renewable
resources. Their properties largely depend on the degree of crystallinity,
crystallization rate, as well as on the size and number of crystallites formed. For
practical applications requiring fast achievement of maximum crystallinity, both
these biobased polymers often need to incorporate nucleation agents. xGnP can
not only act as a nucleating agent, but has the potential to impart higher strength,
modulus, heat resistance, better barrier properties, and electrical conductivity to
polymer matrices.

This study investigates the fundamental interactions between PHB, PLA, and
the graphene surface of xGnP. Once an understanding of how the xGnP affects
the crystallization behavior and nanoscale structure of the polymers is obtained,
insights into how this interaction at the nanoparticle level affects the resulting
nanocomposite properties can be attained. The outcome of the study is of

importance for semicrystalline polymers in general.

Background on nucleation of semicrystalline polymers

One common feature of the thermoplastic polymer matrices used in
nanocomposites, regardless of their production from petroleum or from natural
resources, is their semicrystalline nature. The crystalline state is commonly
defined as one that diffracts X-ray and exhibits a first-order transition known as
meltingf‘2 Upon cooling from the melt state, crystalline polymer crystals nucleate

and grow.

The importance of nucleation for semicrystalline polymers

During melt processing of semicrystalline polymeric materials, depending on
cooling and shear conditions, complex morphologies develop and these
determine the mechanical properties of the materials obtained. For dynamic
applications such as injection molding, faster crystallizing polymers are ideal
candidates.

Certain requirements must be fulfilled in order for a polymer to crystallize.33
The molecular structure of the polymer should be regular enough to allow the
crystalline ordering; the crystallization temperature should be below the melting
temperature, but not too close to the glass transition temperature of the polymer;
nucleation must occur prior to crystallization, and the crystallization rate should
be sufficiently high. Controlling the nucleation is an essential tool for

manipulating the morphology and thus controlling the material properties.

10

A basic classification of the nucleation processes in a polymer melt divides
them into homogeneous, when the melt is free of extrinsic entities, and
heterogeneous, in which case nucleation occurs in the presence of foreign
surfaces in the pure polymer melt. Nanosize reinforcement particles in polymer
composites have been shown to act as heterogeneous nucleating agents in

several semicrystalline matrices.”37

The heterogeneous nucleation of crystalline polymers

Heterogeneous nucleation processes are commonly described in terms of a
free energy balance between the formation of a new crystal and new surface(s)
associated with it.38'39 Initially, there is an increase in free energy associated with
the formation of nuclei smaller than a critical size. Because the newly-formed

crystalline phase is small, the interfacial energy (AGsurface) increases the total

free energy (AG) to a larger extent than the free energy accompanied by the

phase transformation (AGC) lowers it.38 The system remains in

thermodynamically unfavorable state until the free energy reaches a maximum,
and then decreases and becomes negative, leading to stable nuclei
(Figure 1.2)”.

Homogeneous nucleation is rarely observed for most of the crystalline
polymers, since the presence of any foreign substrate (usually impurities,
unmelted polymer, remaining catalysts, or reinforcing particles - in case of

polymer composites) alter the surface energy contribution to the free energy.

11

In order to control the heterogeneous nucleation, one needs to understand
the factors governing the process. The heterogeneous nucleation relies on the
presence of foreign surfaces in the polymer melt (e.g. nucleating agents) that
reduce the free energy opposing primary nucleation (AGH ), increasing the rate
of nucleation, and the rate of crystallization. The free energy opposing primary

nucleation defined38 in Equation 1.2.

 

Unstable
nucleus

0 > Size
Critical \Stable
nucleus

nucleus

 

 

 

 

Figure 1.2 Schematic representation of primary nucleation

=16'A7'7'79'Tri27

AGH I
(AH'AT'PC)2

(1.2)

where A7 is the specific interfacial free energy difference parameter that
accounts for one surface containing the melt and one surface containing the
heterogeneous nucleus, and 7 is its side surface free energy. The most

important parameters affecting the heterogeneous nucleation are supercooling

12

(the difference between the melting temperature and the temperature of

crystallization from the melt), and the specific interfacial free energy parameter.

As the primary nucleation in proportional to 1/AT2 , the higher the supercooling,

the faster the crystallization process takes place. Nucleating agents are more

effective as the specific interfacial free energy parameter, Ay, is smaller. A
smaller A7 ensures a smaller free energy barrier opposing the nucleation

process. Other parameters affecting the heterogeneous nucleation are the
thermal history of the polymer, and epitaxy.38 The thermal history is particularly
important since the presence of non-melted polymer particles that may act as
nuclei is largely depending on melting temperature, melting time, as well as on
the temperature of previous crystallization. Epitaxy, on the other hand, is related
to crystallographic similarities between the topography of the polymer and the
surface topography of nucleating agent, and is still considered the main physical
process explaining the nucleating capability of nucleating agents. This aspect

will be discussed later in this study, with a focus on PHB and PLA.

Nucleating agents for semicrystalline polymers

The addition of nucleating agents to semicrystalline polymers has two
important effects. First, nucleating agents affect the kinetics of crystallization: the
overall rate of crystallization of the polymer increases, leading to faster
solidification of the polymer melt upon cooling, which enhances processability.

Second, the nucleating agents affect the polymer morphology. by diminishing the

13

average spherulite size, which modifies the mechanical and optical properties of
the nucleated polymer.”39 In PNC, the nanosize particles were shown to act as
nucleating agents in various thennoplastic-based systems. The mechanism of
nucleation, and the effects on the polymer properties are largely depending on
the size and nature of the nucleating agents, and are different for nanosize
particles than for the traditional nucleating agents.

Nucleating agents do not affect all semicrystalline polymers to the same
extent. To be sensitive to nucleating agent, a polymer’s crystallization rate has to
be not too high (as, for example, high density polyethylene (HDPE)), in which
case the growth would occur on the formed polymer crystals, regardless of the
presence of the nucleating agent in the melt, but not too low either (e.g.
polycarbonate — an amorphous polymer).

The general requirements for good nucleating agents are: to contain both an
organic and an inorganic group, to be well dispersed in the polymer, to be
insoluble or able to become insoluble in the polymer, to have an epitaxial match
with the polymer crystal, or to chemically react with the polymer generating a
product that can nucleate (in-situ generation of a nucleating agent).38

The nature of nucleating agents can be very diverse. Impurities (such as
catalyst residues), organic compounds, for example benzoic acid, and

40. 41

saccharine , and inorganic compounds, such as talc and pigments, can all

nucleate crystalline polymers.”39
PP, polyamides, PET are extensively modified by using nucleating agents.42

The optimum content of nucleating agents varies for different polymers, but is

14

usually below 1 %, depending on the crystallization behavior of the polymer.
Polypropylenes, for example, have intermediate crystallization rates, and are
being studied to establish structure-properties relationships and to compare the
effects of various nucleating agents. Examples of traditional nucleating agents
for polypropylenes are talc, aluminum salts of dibenzylidene sorbitol
derivatives,“3"“'38 4-biphenyl carboxylic acid, thymine, dibenzyliden sorbitol,

sodium benzoate.38

Efficiency of ngcleating agents

The conventional method used to characterize the efficiency of nucleating
agents is differential scanning calorimetry (DSC). Two parameters determined
using this thermal analysis method are essential for describing the crystallization
in the presence of nucleating agents: the crystallization temperature (Tc), which is
the peak temperature of the crystallization exotherrn, and the crystallization half-
time calculated at a fixed crystallization temperature. Nucleating agents are thus
more efficient if a polymer’s Tc shifts to higher values in their presence, and
respectively if the crystallization half-time is lower for the nucleated polymer than
for the neat polymer. For the purpose of comparing various nucleating agents,

Fillon et al45

defined the nucleation efficiency by establishing a scale for
describing the nucleation ability. The lower limit is defined by the neat, non-
nucleated polymer, while the upper limit is set by the wholly self-nucleated

polymer. The upper limit was set considering that the crystal fragments

produced in self-nucleation satisfy three criteria that characterize an ideal

15

nucleating agent: ideal concentration (i.e. highest achievable), ideal dispersion in
the molten polymer, and ideal polymer-“substrate” interaction (since the polymer
and the “nuclei” have the same chemical composition and crystal lattice). In this

context, the nucleation efficiency (NE) was defined45 as follows (Equation 1.3):

T —T
NE: “NA °1 x100, (1.3)
Tchax 77-01

 

where TQNA is the crystallization temperature of the nucleated polymer; T61 is
the crystallization temperature of the neat polymer, and T02,max is the

crystallization temperature of the self-nucleated polymer.

Effects of ngcleating agents

The addition of nucleating agents to crystalline polymers has two main
effects. The first effect is on the kinetics of crystallization: the rate of overall
crystallization increases. The other effect concerns the polymer morphology, and

consequently the mechanical, optical, and barrier properties.

a) Effects on the kinetics of crystallization. Figure 1.3 shows how the heat of
fusion, directly proportional with the degree of crystallinity of a PP (PP), is higher
for talc-filled PP than for calcium carbonate-filled PP. Also, the heat of fusion

increases as the volume fraction of the filler increases.46

16

 

120 -

115i second heating ’A’
110‘ A~

I
A” \A.'9’ 9/
I O I
I
q-

105 ‘ ,
100 a

\

-‘

   
 

Heat of fusion (J/g)

CO
0|
I

first heating

I I I I

0 0.1 0.2 0.3 0.4
Volume fraction of filler

 

 

(O
O

 

 

 

Figure 1.3 Effect of nucleating agent type and content on the crystallization
temperature of PP matrix. (0) CaCOa; (A) talc

Figure 1.4 shows a comparison of the nucleating effect of four agents on PP,

as well as the dependence of the crystallization temperature on cooling rate.44

 

 

 

 

 

130—
1204
:5
E” 110«
a PP
100. A CC
0 TC
0 BA
0 GA
% 1 I T r 1
0 10 20 30 40 50
Cooling rate (°Clmin)

 

Figure 1.4 Dependence of the crystallization temperature on the cooling rate
for PP, and PP containing CaCOa (CC), talc (TC), p-tert-dibutyl
benzoic acid monohydroxy aluminium (BA), and p-dimethyl-
benzylidene sorbitol (GA)

17

The highest crystallization temperatures for nucleated PP were recorded in
the systems incorporating an aluminum-containing organic compound, followed
by PP nucleated by a small organic molecule, p-dimethyl—benzylidene sorbitol

(GA), and talc-nucleated PP.

b) Effects on morphology. Upon addition of nucleating agents, the size and
number of spherulites decreases, and this affects the mechanical and optical
properties of the nucleated polymer. For example, talc, largely used for the
modification of PP, has platelet geometry, and a strong nucleating effect.“46
The addition of talc particles to molten polypropylenes was shown to increase
mechanical properties such as flexural strength, hardness, and impact strength,
as well as thermal properties (heat distortion temperature). Optical properties
are also affected by the addition of traditional nucleating agents, since the
spherulites formed are smaller in size than in case of the neat polymers. Such
nucleating agents that reduce haze and increase the transparency of

48.42 In

semicrystalline polymers are commonly referred also as clarifying agents
case of polypropylenes, dibenzyliden sorbitol derivatives can be very well
dispersed in the matrix, and have such a nucleating effect. All clarifying agents

are nucleating agents, but not all nucleating agents are clarifying agents

(Figure 1.5).42

18

 

 

 

 

 

 

 

Reference
100 _
I > Increase stiffness Nucleating
_ _ agent
Clarifying
agent
0 l l I
1300 1500 1700 1900 2100
E-Modulus (MPa)

 

 

 

Figure 1.5 Behavior of nucleating and clarifying agents

Pukanzky et al.46 studied the morphology and properties of particulate filled
polymers and attempted to elucidate the extent at which the nucleating effect
affects the polymer properties. This group also underlined the existing
controversy regarding the relationship between aspects of the crystalline
morphology a polymer properties. By comparing PP/Ca003 and PP/talc
systems, they showed that talc, which has a nucleating effect on PP, determines
changes in crystallization and in mechanical properties very different from the
changes induced by calcium carbonate, at the same loading.

Another aspect regarding the effects of nucleating agents on the properties of
the modified polymers is related to the thermal conductivity of the nucleation

L49

additive. Radhakhrisnan et a observed that in case of PP, the higher the

thermal conductivity of the filler, the faster the cooling rate, and the lower the

19

crystallinity value. In the same study, the thermal conductivity of the filler was
reported to affect the skin-core type of morphology of molded products.

For traditional nucleating agents, the extent of the nucleating effect depends
also on the particle size of the nucleating agent. Pukanszki et al.46 showed this
correlation in case of PP/talc systems (Figure 1.6). It is almost unanimously
accepted that the strength and the impact resistance of filled polymers depends
largely on the particle size of the fillers. Smaller particles tend to aggregate more,
which is detrimental for the mechanical properties. Aggregation occurs below a
certain particle size or above a certain specific surface area. The critical size
depends on processing conditions and illustrates the importance of attaining a

high degree of dispersion without aggregation.

 

4o - 0.001 um

Tensile strength

 

 

10 I I I I l
O 0.1 0.2 0.3 0.4 0.5

 

 

 

Volume fraction of filler

 

Figure 1.6 Effect of the size and amount of talc particles on PP

20

In terms of techniques utilized to monitor changes in polymer morphology
determined by nucleating agents, optical microscopy (OM) is a visualization
technique that has been largely used to observe the changes in polymers’
spherulitic structures. AFM is now routinely used to study the surface
morphology and nanostructure of crystalline polymers and polymer

50-5152 offering insights into processes such as crystallization, crystal

composites,
thickening and crystal deformation, phenomena which can also be observed in
situ.5° Changes in crystallinity are usually monitored using WAXD, while small
angle X-ray diffraction methods (SAXD) may be employed to determined

changes in the fine Iamellar structure.

Factors that reduce crystallinity in polymers

External factors such as the nature of solvents or solvent mixtures in case of
crystallization from solution, or temperature, pressure, and crystallization time in
case of crystallization from the melt are known to be controllable to a certain
extent for achieving designed degrees of crystallinity in polymer samples. When
inhibiting the crystallization is a goal, strategies include blending with partially
miscible amorphous polymers”, or incorporating additives with specific surface
active groups that bind to nascent crystals and frustrate crystallization mainly by
acting as defects along the polymer backbone.“ 55

Of particular interest for the present study are geometrical confinement

effects leading to reductions in crystallinity. Observation of crystallization of PET

droplets in an amorphous polycarbonate matrix56 indicated that PET crystallizes

21

differently in bulk than it does when surrounded by the amorphous phase, which
stops crystallization. The finer the dispersion of PET, the greater the inhibition of
crystallization. Similar inhibition of crystallization processes, resulting in a
decrease of the degree of crystallinity, was observed in polymer composites
consisting of alternating layers of two or more polymers (multilayer polymer
films). During the last decade, materials were described where the thickness of
polymer layers varies from micron-size to nanosize.57 The layer thickness affects
the physical properties of the polymer‘r’a'61 and its ability to undergo thermal
transitions. When semicrystalline polymer layers are as thin as a few tens of
nanometers“, confinement of the polymer generates transparent layers, since no
micron-scale structures can be formed.

Nanoscale confinement effects resulting in inhibition of crystallization, as well
as in modification of glass transition temperature, were also observed in PNC, for
example containing OMLSB‘M", or CNT“. In such nanocomposites incorporating
high-modulus, rigid nanoparticles, as the amount of particles in the polymer
matrix increases, the distance between particles decreases, and more polymer
chains are confined. This entrapment was described as polymer sandwiched
between two hard wallsse, and an analogy was made with thin polymer films.
Modeling and understanding of the behavior of confined polymers were more
successful in predicting changes in glass transition temperature and their
dependence on thickness of polymer film/interparticle spacing.67' 68 When the
nanosize particles provide nucleating centers for the polymers, the confinement

leads to inhibition of crystallization. The growth of spherulitic structures is

22

stopped, and bi-dimensional crystalline entities are formed, as seldom proven by

systematic studies of crystallization behavior. Ferreiroag' 7°

reported the formation
of irregular dendritic structures in OMLS/polymer blends and a year later,
Granasy et al.71 described the formation of “dizzy dendrites” using a phase-field
theory, in which randomly distributed foreign particle inclusions perturb the
crystallization by deflecting the tips of the growing dendrite arms. Recently, in a
study regarding HDPE/CNT, Trujillo et al.72 reported that high CNT loadings lead
to a decrease in polymer crystallinity, longer crystallization time, and sharp
decrease of Avrami exponent for isothermal crystallization. The Avrami exponent
calculated for low amounts of CNT in HDPE was between 1.94 and 2.2,
corresponding to three-dimensional spherulitic structures, and dropped
dramatically as the CNT loading was increased. Such low values are commonly

attributed to one-dimensional and two-dimensional crystal structures. Similar

effects are described further in this thesis (Chapters 3 and 5).

Nanoparticles nucleating semicrystalline polymers

The traditional nucleating agents used for the modification of semicrystalline
polymers are usually micron-sized. As PNC research developed, numerous
nanoparticles used as reinforcements in polymer matrices were shown to act as
heterogeneous nucleating agents for the polymers. In essence, both nano— and
micron-size particles are heterogeneous nucleating agents. The main difference

is due to the size effect, largely discussed in relation to the mechanical properties

23

of polymer composites, which plays a significant role in the early stages of
nucleation, since it is related to important modifications of morphology.

It has been shown35‘34' 73‘" that nanoparticles can be used as structure and
morphology directors, for example for stabilizing a metastable or conventionally
inaccessible polymer phase”, or to induce new energy dissipation
mechanisms.76

In terms of processing, it is important to notice that, as opposed to phase
separation — occurring in case of incorporating micro-sized fillers in polymers -,
as the size of the fillers in the polymer matrices decreases, dispersion and
agglomeration become key problems to be overcome in obtaining
high-performance polymer composites.

Polymer materials containing nanoparticles have different failure
mechanisms. As the size of the fillers decreases, the materials no longer fail due
to debonding, which is the deformation mechanism predominating in composites
containing large particles; cracks are initiated inside aggregates forming upon
addition of particles having small size.77

A novel application of the nucleating effect of nanoparticles (namely OMLS)
on polymer blends was reported recently by Si et al.78 OMLS were shown to act
as compatibilizers in polymer blends, due to their extremely large surface area
and to their ability to become very effective nucleating agents for in-situ graft
formation during melt mixing. An increased compatibility due to the localization

of OMLS particles at the interface was demonstrated for polystyrene/poly(methyl

24

metacrylate) blends, reflected in the increase of the modulus of the blend, and
easier recyclability.

According to Ray, et al.79

, the polymers’ biodegradability can be affected by
the incorporation of nanoparticles. Organically modified layered silicates
(OMLS)/PLA were shown to degrade faster when the OMLS were well-distributed
in the polymer matrix. The authors hypothesized that as maximum parts of the
matrix are in contact with the clay's edge and surface, there is a higher tendency
of the polymer chains to fragment than in case of poor dispersion of the
nanoparticles. To our knowledge, correlations between the rate of
biodegradability and the dispersion of micro-sized fillers and/or nucleating agents
were not reported. The potential effect on the rate of degradation is another
unique feature of nanosize materials, again due to the size similarity between the
nanoparticles and the polymers’ nanoscale structure.

Besides the OMLS, that were proved to nucleate semicrystalline polymers,
and to promote different polymorphs, affecting the properties of the resulted
nanocomposites, another significant class of nanosize particles largely
investigated is represented by CNTs. Various forms of CNT were shown to
nucleate crystalline polymers. Several research groups investigated the
nucleating effect of single-walled carbon nanotubes (SWNT) on PPM}55 Grady et
al. showed that the SWNT changed the kinetics of both isothermal and
nonisothermal crystallization, in a manner similar to carbon fibers. The amount

of SWNT required to nucleate PP was found80 to be 0.6 % wt, at which loading

the number of nucleation sites increased up to the point that more SWNT did no

25

longer increase the number of crystallites. The very low amount confirmed that
the number of nucleated crystals was proportional to the surface area of the
nucleating agent. Recently, CNT were showed to nucleate poly(hydroxyvalerate-
co-hydroxybutyrate)“, PVABZ, PLA3‘, PET83, HDPE“, Nylon-685, Polyamide

101085.

Nanocomposites based on biodegradable polymers

Biodegradable polymers are defined as polymers that undergo microbially
induced chain scission leading to mineralization.87 Such polymers can be derived
from petroleum sources or can be obtained from mixed petroleum and
biomass-derived products. A classification of biodegradable polymers is
presented in Table 1.1.

Biodegradable polymers obtained from biologically based feedstock are
especially attractive since these reduce the reliance on polymers derived from
depleting natural resources. In the fast-developing research on PNC, during the
last ten years there has been an intensive research on biodegradable polymers

as matrices, especially containing OMLS reinforcements.

26

Table 1.1 Classification of biodegradable polymers

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Natural polymers Synthetic polymers
Sub-classification Examples Sub-classification Examples
1. Plant origin 1. Aliphatic polyesters
1.1. Polysaccharides cellulose, starch, 1.1. Glycol and poly(ethene succinate),
alginate dicarbonic acid poly(butylenes
polycondensates terephthalate)
2. Animal origin 1.2. Polylactides polyglycolide,
polylactides
2.1. Polysaccharides chitin (chitosan), 1.3. Polylactones poly(s-caprolactone)
hyaluronate
2.2. Proteins collagen, albumin 1.4. Miscellaneous poly(butylenes
terephthalate)
3. Microbial origin 2. Polyols poly(vinyl alcohol)
3.1. Polyesters poly(hydroxyalkanoates) 3. Polycarbonates poly(ester carbonate)
3.2. Polysaccharides 4. Miscellaneous polyanhydrides,
poly(a—cyanoacrylate)s,
polyphosphazenes,
poly(orthoesters)

 

 

 

 

Poly(hydroxybutyrate) and polylactic acid polymers

One group of biodegradable polymers that has received considerable
attention is the poly(hydroxyalkanoates) (PHAs),89' 9° which are thermoplastic
biopolyesters typically synthesized by various bacteria and microorganisms that
use them as reserves of carbon and energy.8H1 Materials produced from PHAs

can be totally biodegraded under aerobic and/or anaerobic conditions 8994 at the

27

 

end of their lifetimes. Figure 1.7 shows the chemical structure of PHAs. In PHB,

the side group is CH3.

 

 

Slde group

j
Hydroxyacld Ho - CH - CH2— COOH

 

 

 

I313

‘7"
RIZATIDN

 

 

 

 

 

 

 

 

;
Poumg
V
Poly-hydmxyacld
Side group 0 Slde group O
I II I ll
O—CH—CHz-C O-CH-CHz—C

etc.

 

 

 

Figure 1.7 Chemical structure of PHAs

Polylactic acid or polylactide is a biodegradable, thermoplastic, aliphatic
polyester derived from lactic acid. PLA is commonly produced in a high
molecular weight by ring-Opening polymerization of lactide (Figure 1.8).

Several forms of PLA exist, and the variety is due to the chirality of the lactic
acid. Poly-L-lactide (PLLA) is the product resulting from polymerization of lactid
acid in the L form. PLLA has a crystallinity around 37 %, a glass transition
temperature between 50 and 80° C and a melting temperature between 173 and
178° C. The polymerization of a mixture of both L and D forms of lactic acid leads
to the synthesis of poly-DL-lactide (PDLLA) which is not crystalline but

amorphous.

28

 

 

 

catE'ySt r CH 0 \ CH 0
H3C 0 Heat I3 II I3 ||
‘ —-O—CH—C-—O-CH-C—
0 CH3 & J n
O
Lactide Polylactide

 

 

 

Figure 1.8 Synthesis of polylactides

PHB and PLA are two of the most investigated biodegradable thermoplastic
polymers derived from natural resources. They are both semicrystalline in nature,
and can be processed using conventional polymer processing methods, where it
is desired that the polymers crystallize fast from the melt. One way to ensure
fast crystallization is to incorporate nucleating agents or reinforcements that
could help control the crystallization rate. Similar to traditional thermoplastic, the
expected effects after adding nucleation agents are: improvement of processing
cycle or spinnability, improvement in transparency of molded articles,
improvement of rigidity of molded articles or filaments, improvement of

dimensional precision, improvement of gloss, and prevention of sink.43

29

The importance of nucleation for PHB and PLA

a) Nucleation of poly(hydroxybutyrate)

PHB, in addition to being biodegradable and biocompatible, is attractive due
to its availability, processibility, and mechanical and barrier properties which are
comparable to isotactic PP and other synthetic polymers.95' 96 PHB is also a
highly crystalline thermoplastic that can be extruded, injection molded, and

97' 98 into fibers without modification of traditional polymer processing

spun
equipment. PHB also has a low elongation at break (less than 10 %), an impact
strength of 3 J/mmz, a high modulus (1.7 GPa) and a fracture stress of 35 MPa.99
Despite having these desirable characteristics, PHB is not widely applied in the
manufacturing industry, primarily due to its narrow processibility window (PHB
melts around 180 °C,92 very close to its thermal degradation temperature), and to
its brittleness which is related to its crystallization behavior. A host of
researchers has attempted to resolve these processability issues by

100. 101

co-polymerizing PHB with other hydroxyalkanoates,91' mainly

hydroxyvalerate and hydroxyhexanoate, blending with poly(ethylene oxide), PVA,

102-105.105,106,107

or poly(lactide) for example. adding plasticizers like Citric esters,108

109

and processing aids, and/or through annealing.110

The crystallographic structure of PHB was elucidated in the early 1970311"112
PHB crystallizes in an orthorhombic system as 21 chains aligned in groups within

an amorphous mass. The lattice parameters are

0
a = 5.76, b = 13.20 and C = 5.96A, with 0 corresponding to the fiber period,

30

R = 0.13. The chains are stacked in well-aligned rows giving two-dimensional
order, but the polymer is ill defined along the third axis113. Gazzano et al.114
reported the XRD diffraction patterns for PHB powder, as well as for PHB
isotherrnally crystallized at different temperatures. The changes with varying the
crystallization temperature were attributed to strong orientation effects associated
with the morphological variations induced by different temperatures. The lattice
parameters were not found to change significantly, but the size of the sphemlites
formed at different Tc varied.

PHB crystallizes forming large banded spherulites, the structure of which has
largely been investigated. The most recent article, published by Tanaka et al.115
in 2005, gives more insight on the structure of the banded spherulites.

The brittleness of PHB1°9'"6'"7'"3'119 is known to stem from three factors: a
glass transition temperature close to room temperature, secondary crystallization
occurring upon storage at room temperature, and extremely low nucleation
density. All these factors are interrelated, and derive from the high purity of PHB

50,116

and its stereochemical regularity. Detailed studies of the crystallization

3120, outline the importance of physical

phenomena occurring in bacterial PH
aging as an intrinsic property of this polymer, attributed to progressive
crystallization occurring upon storage at room temperature. As a result of its
biosynthetic preparation, solvent-based extraction and purification processes,
PHB is exceptionally pure and has a very low level of heterogeneous nuclei that

could initiate crystallization.118 Due to its low nucleation density, when cooled

from the melt state, PHB forms large spherulitic structures that continue growing

31

into each other during storage, and are prone to cracking, resulting in materials
with low impact resistance and high brittleness.

There is an ongoing research thrust to discover efficient nucleating agents for
PHB and its copolymers. Efficient nucleating agents would increase the
polymer’s crystallization temperature, increase the crystallization rates, and
generate smaller and more numerous spherulites, leading to materials with
increased mechanical properties. Compounds ranging from talc,m'122'123 boron
nitride,122'123'4° terbium oxide,123 lanthanum oxide,123 saccharin,“"'41 phtalimide,4o
cyclodextrin,121 and lignin,124 were investigated as possible nucleating agents for
PHB and several nucleation mechanisms were proposed.122""'12“'118 Among the
nucleating agents mentioned above, micro- and nanosize inorganic particles
were shown to have the added advantage of acting as reinforcements for the
polymer. For example, OMLS nanoplatelets, commonly used as reinforcements

125126 were proven to efficiently nucleate and to

for thermoplastic polymers,
reinforce PHB36 and poly(3-hydroxybutyrate-co-3-hydroxyvalerate).109 Also,
single-walled carbon nanotubes were shown to nucleate PP80 and to affect the
thermal properties and the morphology of poly(hydroxybutyrate-CD-
hydroxyvalerate).81

However, what is apparent from all these studies is that the Choice of
nucleating agent should be based upon an understanding of the mechanism by

which nucleation occurs. Barham118 summarized the different types of nucleating

agents as follows:

32

Type A: nucleation from the pure melt This is a slow nucleation which occurs
when PHB is cooled from a melt or above ~210 °C, and does not occur at
predetermined sites.

Type B: self-seeding: This occurs when PHB is heated no more than ~5 °C
above its melting point; there is a higher nucleation density for lower
superheatings, and no increased nucleation density if the melt temperature
exceeds 205 °C or if the superheating exceeds 200 °C. In addition, the nucleation
is reduced if the melt is held at high temperature for more than a few minutes.

Type C: enhanced self-nucleation: This is an extension of self—seeding.
When the samples are heated to a temperature exceeding the melting point with
maximum 1°C, and then cooled to a crystallization temperature, spherulites
simply reappear, as a reverse of melting, not by the usual radial growth process.

Type D: nucleation by impurities: For example, talc (5 pm) may act as an
impurity and stabilize the self-seeding type of nucleation. The authors of this
Classification of nucleation mechanisms for PHB specified that if the impurity is
added directly to the melt, it will have no nucleating effect; it is only after
crystallization of PHB once occurred around the talc that the talc acted as
nucleating agent after subsequent melting.

Type E: “classical nucleation”: Nucleation by saccharin (particle diameter
approximately 50 pm) shows this type of behavior, and occurs both when the
agent is added directly to the melt, and when the polymer is first allowed to
crystallize around it. Nucleation occurs preferentially on certain faces; it starts on

111 faces, above which the oxygen atoms project; oxygen atoms are spread

33

approximately 1.185 nm apart. PHB has an axial repeat of ~0.596 nm and fits
with little distortion on the saccharin surface.

Barhamm'118 also proposed a theoretical interpretation of the nucleation rate
data collected. The work required for forming a nucleus, in terms of surface free

energies and the bulk free energy of fusion, as a function of crystal dimensions,

*

was calculated. The rate of nucleation is proportional toexp[— 1:; ], where

 

A¢' is the work to form the critical nucleus which, for homogeneous nucleation is
given by Equation 1.4.
. _ 320206

A ——, 1.4
¢ (AGV)2 ( )

where a is the side surface energy. 0e is the end, or fold, surface free energy,

and AGV is the bulk free energy of fusion. The shape of the nucleus was

assumed to be a rectangular parallelepiped. For the heterogeneous nucleation
on a foreign surface, all the three dimensions of the nucleus are allowed to
Change. In this case, the work required to form a critical nucleus given by

Equation 1.5.

. _ 160209130

A _ 1.5
¢ (AGV)2 ( )

 

where Aa=a+croS —as, 0s is the interfacial energy between the foreign

surface and the polymer melt, and 0'03 is the interfacial energy between the

34

polymer crystal and the foreign surface. In the absence of a substrate, as = 0
and 00s = 0', hence A0 = 20. The authors outlined that it was important for the
foreign surface to resemble the crystal surface; Us is closer to a; Gas is closer

to zero, and A0 is smaller. The extreme case is obtained when the crystal is

identical to the foreign surface, and A0 = 0. Then the work is then:

r _ 4b0'0',9

13¢ AGV .

 

(1.6)

where b is the lattice spacing normal to the surface. The bulk free energy of

fusion is:
AGV = AH — TAS. (1.7)
At T = T3 (equilibrium melting point), AGV = 0, so:

AHAT
0
TM

 

AGV = — , (1.8)
where AH is the enthalpy of fusion, and AT is the supercooling.

A different behavior was observed for the PHB/talc systems, in which the
PHB molecules either formed a crystalline layer around the talc, which prevented
the polymer from melting, leaving a polymer crystal substrate for subsequent
nucleation, or the PHB molecules were partially absorbed onto the talc particles,

which reduced their entropy in the molten state. The nucleating effect of

35

saccharin was explained by the absorption of molecules onto the surface in what
is close to their crystallographic form.

Another study reported by the Barham117 refers to the crystallization and
morphology PHB. The main purpose of the study was to introduce PHB as a
model polymer for studies of morphology and crystallization, and to determine
the heat of fusion, surface free energies, and other constants important for
further kinetics studies. The effect of seeding on the crystallization of
poly(hydroxybutyrate), and co-poly(hydroxybutyrate-co-valerate) was also
explored by Withey et al.40 The enhanced self-seeding of PHB upon addition of
PHB crystals was also investigated. Additionally, mixtures of PHB with saccharin
and phtalimide were studied in the view of explaining the nucleating effects. The
development of crystallinity within the mixtures, and the relative ability, and the
relative ability of the nucleating agents to induce crystallization of PHB/PHBV
were studied. Samples were prepared by compression molding, and analyzed
using polarized optical microscopy (with a hot stage), SEM imaging of the
fracture surfaces, FT-IR, and DSC. Boron nitride was found to nucleate better
than saccharin and phtalimide. Phtalimide had almost no nucleating ability, and
both saccharin and phtalimide appeared to be soluble in the melt but were less
soluble in the crystalline regions. These findings are unique since FT-IR was
used for microstructural analysis, in order to establish local compositions within
the boundaries of the spherulites. It was pointed out that saccharin and
phtalimide were excluded from the growing polymer crystallites and accumulated

at the interface between the spherulites boundary and the melt. The conclusion

36

was that the mixtures formed a complex-phase system with both saccharin and
phtalimide being present at the intersection of polymer spherulites.

Recently, Kai et al.122

reported a Closer look at the fast crystallization of
poly(3-hydroxybutyrate) and poly(3-hydroxybutyrate-co-3-hydroxyvalerate) with
talc and boron nitride as nucleating agents. Two nucleation mechanisms were

proposed for boron nitride and talc.

b) Nucleation of polylactic acid

The crystallization rates of various types of poly(lactic acid) are very slow as
compared with commercially useful therrnoplastics such as polyethylene, and it
has been established that there is a clear need to use nucleating agents to
modify this polymer’s crystallization behavior. The traditional approach of
employing heterogeneous nucleating agents in order to reduce process cycle
times, increase the crystallization temperature, and diminish the size of

crystallites, was investigated for PLA by numerous authorsm'128

As pointed out by Schmidt et al. ‘28

, without the addition of nucleating agent,
the half-time of crystallization (t1/2) is very high, 1.9 minutes for a PPLA sample
with a molecular weight of ~100 kg/mol. For a similar PLLA samplem, 6 wt %
talc, a common nucleating agent for semicrystalline polymers, was found to
increase the crystallization rate of PLLA. In this case, t1/2 was less than one
minute at crystallization temperatures between 90 °C and 120 °C.

Organically modified layered silicate nanocomposites (OMLS) of PLA were

first reported in 1997 by Ogata et al37. The PLA/OMLS were prepared by

37

dissolving the PLA in hot chloroform in the presence of a modified
montmorillonite (MMT), and showed an increase in Young’s modulus as
compared to the neat PLA. In this very first study, the layered silicates were
shown to act as nucleating agents for PLA. Figure 1.9 shows the DSC
therrnograms of PLLA/OMLS blends”. The exothermic peaks in the temperature
region between the glass transition temperature and the melting temperature
correspond to cold crystallization. The peaks are narrower for PLA/OMLS blends
than for PLA, and the fact that the peak temperatures are lower for PLA
containing OMLS than for the neat PLA suggests that the small OMLS particles

act as nucleating agents.

 

20%
10%
5%

3%
0%

Endo H Exo

 

 

50 100 150 200 250 300
Temperature (°C)

 

 

 

Figure 1.9 DSC curves of quenched PLLA/OMLS blends containing different
amounts of OMLS

38

Extensive research was being conducted on PLA/layered silicates
nanocomposites by the Okamoto group at the Toyota Technological Institute in
Japan. The same nucleating effect was observed in nanocomposites obtained

from organically modified synthetic fluorine mica (OMSFM)129. Figure 1.1

0130
shows the DSC curves obtained during heating of PLA, and PLA containing 4 wt
% OMSFM, and the polarized optical micrographs obtained for the same
samples. The neat PLA crystallized at 110 °C had a well-defined, large
spherulitic morphology, while the size of the spherulites decreased upon the
addition of the OMSFM, which evidently increased the number of nuclei and
modified PLA’s crystallization behavior.

The same group reported on the effect of layered silicates on the
crystallization kinetics of PLA‘31'132, underlining the importance of knowing the
crystallization of PLA in detail, and how it is affected by the presence of OMLS
nanoplatelets. In their first study of the crystallization behavior of PLA/clay
systems, crystallization kinetics of nanocomposites obtained from PLA and
octadecylammonium modified montmorillonite prepared by extrusion was
investigated using time-resolved LS photometry. The dependence of the overall
crystallization rate (Figure 2.10), and of the diameter of PLA spherulites on the
crystallization rate was rigorously quantified. The kinetic analysis, together with a

polarized optical microscopy study, revealed that the dispersed clay particles

acted as nucleating agents for PLA.

39

 

U
v

2 — Neat PLA

Heat Flow (mW)
O

_4_ PLACN4

.6 I I I I I 1
-40 0 40 80120160200

Temperature (°C)

 

 

b)

Neat PLA

 

PLACN4

 

 

 

 

Figure 1.10 DSC traces of neat PLA and PLNCN. (b) Polarized optical

micrographs of crystallized (at 110 °C) samples of neat PLA and
PLA/CN

Moreover, for PLA and PLA/OMLS, the group reported on the dependence of
the spherulite radius, and of the overall crystallization rate on the crystallization

temperature131 (Figure 1.11), as well as on the relationship between the

nucleation density and the crystallization temperature.

40

 

1-OE-01 i . PLA CN4
o Neat PLA
I ) ' 3
d Q6 [030 1.0E—02 ‘ - O (.3
dt 0
1.0E-03 — o '
u o
8
1.0E-04 r I

 

 

70 80 90 100 110 120 130 140

 

 

Crystallization temperature (°C)

 

Figure 1.11 Dependence of the overall crystallization rate of PLA, and
PLA/OMLS nanocomposites on the crystallization temperature

A subsequent study compared five different organically modified layered
silicates in terms of their effects on the crystallization of PLA. The conclusion
was that the surface of OMLS acts as nucleating agent, increasing the
crystallization temperature, and affecting the morphology of the resulted
nanocomposites. The PLNOMLS systems were proven to undergo
heterogeneous nucleation kinetics. An important observation regarding
morphology was that the spherulites of PLA/OMLS had a less ordered structure
than those of pure PLA, due to the OMLS particles that were not nucleating and
were dispersed in the spherulites, disturbing the regular orientation of the Iamella

stacks inside the spherulites.131

41

Exfoliated graphite nanoplatelets as nucleating agents

The structure of graphite is very similar to the one of boron nitride

(Figure 1.12), which was previously shown to act as efficient nucleating agent for

133

PHB. xGnP were also shown to act as nucleating agents and reinforcement

1333029

for other thermoplastic polymers , and have a series of extremely structure-

related attractive advantages.

 

 

  

o Boron atom
o Nitrogen atom

 

 

 

Figure 1.12 Boron nitride crystal structures

Graphite is a layered natural mineral, and the layers can be intercalated and
exfoliated into nanosize platelets with high aspect ratio, high surface area and
in-plane stiffness as high as 1060 GPa.134 While the basal plane is inert
(consisting of carbon in its sp2 state along with a graphene surface rich in 1t
electrons), the existence of functional groups (Figure 1.13) at the edges can lead
to hydrogen or covalent bond with polymer matrices. However, it should be
noted that with the typical large aspect ratios of xGnP (~100) the amount of edge

sites compared to basal plane sites is very small.

42

 

Carboxyl

       
   

0 Lactone
Pyrone

 

\NHz \\NH 0 Carbonyl

Amine Imine

 

 

 

Figure 1.13 Schematic of the xGnP

Exfoliated graphite nanoplatelets having an average size of 1 pm (xGnP-1)
were produced in—houseze' 135 starting from GrafGardTM 160-50A acid intercalated
graphite produced by Graftech (Cleveland, OH). The xGnP prepared with our
MSU method has a surface area of approximately 100 m2/g and consists of
platelets approximately 10 nm in thickness and 1 pm in diameter. The platelet
basal plane is a graphene sheet, uniform in structure and surface energy, with
evidence of some oxygen functionalities located at the edges of these basal
surfaces.

Recent research133 in the Drzal group showed that the graphene basal plane
promotes nucleation of PP. At low xGnP concentration, PP was shown to
crystallize in a less common ,B-form, consisting in crystals with higher impact
strength. Besides the concentration of xGnP, crystallization conditions were also

shown to affect the population and size of PP spherulites, which clearly

43

influenced the electrical conductivity of the resulted PP/xGnP nanocomposites.
The PPIxGnP study indicates that smaller and more numerous crystallites
formed by PP in the presence of xGnP may disrupt the formation of conductive
network, thus increasing the percolation threshold for electrical conductivity. This
research strongly indicates that in order to achieve control over nanocomposite
properties it is extremely important to understand the effects of xGnP on the

crystallization of the polymer matrix in which the nanoplatelets are incorporated.

44

References

10.

11.

12.

13.

Okada, A.; Usuki, A., Twenty years of polymer-clay nanocomposites.
Macromolecular Materials and Engineering 2006, 291, (12), 1449.

Krishnamoorti, R.; Vaia, R. A., Polymer nanocomposites. Journal of
Polymer Science Part B-Polymer Physics 2007, 45, (24), 3252.

Winey, K. l.; Vaia, R. A., Polymer nanocomposites. MRS Bulletin 2007,
32, (4), 314.

The Freedonia Group Inc., Cleveland, OH, Nanocomposites, study
published March 1, 2006 (www. marketresearch.com), accesses May 20,
2006)

Anastas, T. P. W., J.C., Green Chemistry: Theory and Practice. 1st Ed.,
Oxford University Press, New York, 1998.

Hunter, D. L.; Kamena, K. W.; Paul, D. R., Processing and properties of
polymers modified by clays. MRS Bulletin 2007, 32, (4), 323.

Rosato, 0., Plastic nanocomposites technology and trends,
www.specialchem4polvmers.com, SpecialChem SA 2007 (accessed
January 10, 2008).

Vaia, R. A.; Maguire, J. F., Polymer nanocomposites with prescribed
morphology: Going beyond nanoparticle-filled polymers. Chemistry of
Materials 2007, 19, (11), 2736.

Fukushima, Y.; lnagaki, 8., Synthesis of an lntercalated Compound of
Montmorillonite and 6-Polyamide. Journal of Inclusion Phenomena 1987,
5, (4), 473.

Usuki, A.; Hasegawa, N.; Kato, M., Polymer-Clay nanocomposites.
Inorganic Polymeric Nanocomposites and Membranes, 2005, 179, 135.

Usuki, A.; Kojima, Y.; Kawasumi, M.; Okada, A.; Fukushima, Y.; Kurauchi,
T.; Kamigaito, 0., Synthesis of Nylon 6-Clay Hybrid. Journal of Materials
Research 1993, 8, (5), 1179.

Lincoln, D. M.; Vaia, R. A.; Wang, Z. G.; Hsiao, B. 8., Secondary structure
and elevated temperature crystallite morphology of nylon-6/layered silicate
nanocomposites. Polymer 2001, 42, (4), 1621.

Van der Hart, D. L.; Asano, A.; Gilman, J. W., Solid-state NMR
investigation of paramagnetic nylon-6 Clay nanocomposites.

45

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

1. Crystallinity, morphologY. and the direct influence of Fe3+ on nuclear
spins. Chemistry of Materials 2001 , 13, (10), 3781.

Han, B.; Ji, G. D.; Wu, S. S.; Shen, J., Preparation and characterization of
nylon 66/montmorillonite nanocomposites with co-treated
montmorillonites. European Polymer Journal 2003, 39, (8), 1641.

Liu, X. H.; Wu, 0. J., Phase transition in nylon 6/clay nanocomposites on
annealing. Polymer 2002, 43, (6), 1933.

Liu, X. H.; Wu, Q. J.; Berglund, L. A., Polymorphism in polyamide 66/clay
nanocomposites. Polymer 2002, 43, (18), 4967.

Liu, X. H.; Wu, Q. J.; Berglund, L. A.; Di, 2. N., Investigation on unusual
crystallization behavior in polyamide 6/montmorillonite nanocomposites.
Macromolecular Materials and Engineering 2002, 287, (8), 515.

Tseng, C. R.; Wu, S. C.; Wu, J. J.; Chang, F. C., Crystallization behavior
of syndiotactic polystyrene nanocomposites for melt- and cold-
crystallizations. Journal of Applied Polymer Science 2002, 86, (10), 2492.

Strawhecker, K. E.; Manias, E., Structure and properties of poly(vinyl
alcohol)/Na+ montmorillonite nanocomposites. Chemistry of Materials
2000, 12, (10), 2943.

Nam, P. H.; Maiti, P.; Okamoto, M.; Kotaka, T.; Hasegawa, N.; Usuki, A.,
A hierarchical structure and properties of intercalated polypropylene/Clay
nanocomposites. Polymer 2001,42, (23), 9633.

Park, S. Y.; Cho, Y. H.; Vaia, R. A., Three-dimensional structure of the
zone-drawn film of the nylon-6/Iayered silicate nanocomposites.
Macromolecules 2005, 38, (5), 1729.

Schadler, L. S.; Kumar, S. K.; Benicewicz, B. 0.; Lewis, S. L.; Harton, S.
E., Designed interfaces in polymer nanocomposites: A fundamental
viewpoint. MRS Bulletin 2007, 32, (4), 335.

US_EPA Facts. http://www.epa.gov/epaoswer/non-hw/muncpI/index.htm
(accessed February 27, 2006).

US_EPA Recycle. http://www.epa.gov/epaoswer/non-
hw/muncpl/recycle.htm (accessed February 27, 2006).

Joint Services Pollution Prevention Opportunity Handbook.
http://p2library.nfesc.navy.mil/P2_Opportunity_Handbook/7_I_A_7.htmI
(accessed February 10, 2006).

46

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

Fukushima, H. Graphite nanoreinforcements in polymer nanocomposites.
Ph.D. Dissertation, Michigan State University, East Lansing, 2003.

Kalaitzidou, K.; Fukushima, H.; Drzal, L. T., Multifunctional polypropylene
composites produced by incorporation of exfoliated graphite nanoplatelets.
Carbon 2007, 45, (7), 1446.

Kalaitzidou, K.; Fukushima, H.; Drzal, L. T., A new compounding method
for exfoliated graphite-polypropylene nanocomposites with enhanced
flexural properties and lower percolation threshold. Composites Science
and Technology 2007, 67, (10), 2045.

Kalaitzidou, K.; Fukushima, H.; Drzal, L. T., Mechanical properties and
morphological Characterization of exfoliated graphite-polypropylene
nanocomposites. Composites Part A-Applied Science and Manufacturing
2007, 38, (7), 1675.

Kalaitzidou, K.; Fukushima, H.; Miyagawa, H.; Drzal, L. T., Flexural and
tensile moduli of polypropylene nanocomposites and comparison of
experrnental data to Halpin-Tsai and Tendon-Wang models. Polymer
Engineering and Science 2007, 47, (11), 1796.

Supronowicz, P. R.; Ajayan, P. M.; Ullmann, K. R.; Arulanandam, B. P.;
Metzger, D. W.; Bizios, R., Novel current-conducting composite substrates
for exposing osteoblasts to alternating current stimulation. Journal of
Biomedical Materials Research 2002, 59, (3), 499.

Sperling, L. H., Introduction to Physical Polymer Science. 3rd Ed., John
Whiley & Sons Inc. 2001.

Jansen, J., Gachter and Miiller, in Plastics Additives Handbook, Hanser
Publishers,: Munich 1990.

Shah, D.; Maiti, P.; Gunn, E.; Schmidt, D. F.; Jiang, D. D.; Batt, C. A.;
Giannelis, E. R., Dramatic enhancements in toughness of polyvinylidene
fluoride nanocomposites via nanoclay-directed crystal structure and
morphology. Advanced Materials 2004, 16, (14), 1173.

Maiti, P.; Nam, P. H.; Okamoto, M.; Hasegawa, N.; Usuki, A., Influence of
crystallization on intercalation, morphology, and mechanical properties of
polypropylene/clay nanocomposites. Macromolecules 2002, 35, (6), 2042.

Maiti, P.; Batt, C. A.; Giannelis, E. P., Renewable plastics: Synthesis and
properties of PHB nanocomposites. Abstracts of Papers of the American
Chemical Society 2003, 225, U665.

47

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

Ogata, N.; Jimenez, G.; Kawai, H.; Ogihara, T., Structure and
thermal/mechanical properties of poly(l-Iactide)-Clay blend. Journal of
Polymer Science Part B-Polymer Physics 1997, 35, (2), 389.

Kurja, J., Mehl, NA, in Plastics Additives Handbook, 5th Ed., Hanser
Publishers: Munich, 2000.

Zhou, H., Nucleation and nucleating agents in crystalline polymers: a
literature review, in Technical Information Series, GE Research &
Development Center: 1998.

Withey, R. E.; Hay, J. N., The effect of seeding on the crystallisation of
poly(hydroxybutyrate), and co-poly(hydroxybutyrate-co—valerate). Polymer
1999, 40, (18), 5147.

Black, S. N.; Dobbs, B.; Dempsey, P. S.; Davey, R. J., Crystal-chemistry
and nucleating agents - saccharin and poly-3-hydroxybutyrate. Journal of
Materials Science Letters 1990, 9, (1), 51.

Hoffmann, K.; Huber, G.; Mader, D., Nucleating and Clarifying agents for
polyolefins. Macromolecular Symposia 2001, 176, 83.

Fujiyama, M.; Wakino, T., Structures and Properties of Injection Moldings
of Crystallization Nucleator—Added Polypropylenes .1. Structure Property
Relationships. Journal of Applied Polymer Science 1991, 42, (10), 2739.

Fujiyama, M.; Wakino, T., Distribution of Higher-Order Structures in
Injection Moldings of Particulate-Filled Polypropylenes. Journal of Applied
Polymer Science 1991 , 43, (1), 97.

Fillon, B.; Lotz, B.; Thierry, A.; Wittmann, J. C., Self-Nucleation and
Enhanced Nucleation of Polymers - Definition of a Convenient
Calorimetric Efficiency Scale and Evaluation of Nucleating Additives in
Isotactic Polypropylene (Alpha—Phase). Journal of Polymer Science Part
B-Polymer Physics 1993, 31 , (10), 1395.

Pukanszky, B.; Moczo, J., Morphology and properties of particulate filled
polymers. Macromolecular Symposia 2004, 214, 115.

Varga, J.; Toth, F. 8., Filled Compounds of the Beta-Modification of
Polypropylene. Angewandte Makromolekulare Chemie 1991, 188, 11.

Wang, H.; Li, C. C.; Ke, Y. C.; Zhang, D.; Li, Z. Y., Effects of rosin-type
clarifying agent on the crystallization and compatibility of polypropylene
and low density polyethylene. Journal of Applied Polymer Science 2006,
99, (4), 1568.

48

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

Radhakrishnan, S.; Sonawane, P.; Pawaskar, N., Effect of thermal
conductivity and heat transfer on crystallization, structure, and morphology
of polypropylene containing different fillers. Journal of Applied Polymer
Science 2004, 93, (2), 615.

Hobbs, J. K.; Winkel, A. K.; McMaster, T. J.; Humphris, A. D. L.; Baker, A.
A.; Blakely, S.; Aissaoui, M.; Miles, M. J., Some recent developments in
SPM of crystalline polymers. Macromolecular Symposia 2001, 167, 1.

Binnig, G.; Rohrer, H., In touch with atoms. Reviews of Modern Physics
1999, 71, (2), S324.

Quate, C. F., Vacuum Tunneling - a new technique for microscopy.
Physics Today 1986, 39, (8), 26.

Wu, D. F.; Zhang, Y. S.; Zhang, M.; Wu, L. F., Morphology, nonisothermal
crystallization behavior, and kinetics of poly(phenylene
sulfide)lpolycarbonate blend. Journal of Applied Polymer Science 2007,
105, (2), 739.

Stowe, M. K.; Liu, R; Baker, G. L., Star poly(ethylene oxide) as a low
temperature electrolyte and crystallization inhibitor. Chemistry of Materials
2005, 17, (26), 6555.

Lopez, L. C.; Wilkes, G. L.; Geibel, J. F., Crystallization Kinetics of
Poly(Para-Phenylene Sulfide) - the Effect of Branching Agent Content and
Endgroup Counter-Atom. Polymer 1989, 30, (1), 147.

Molinuevo, C. H.; Mendez, G. A.; Muller, A. J., Nucleation and
crystallization of PET droplets dispersed in an amorphous PC matrix.
Journal of Applied Polymer Science 1998, 70, (9), 1725.

Adhikari, R; Lebek, W.; Godehardt, R.; Henning, 8.; Michler, G. H.; Baer,
E.; Hiltner, A., Investigating morphology and deformation behavior of
multilayered PC/PET composites. Polymers for Advanced Technologies
2005, 16, (2-3), 95.

Balsamo, V.; Stadler, R., Influence of the crystallization temperature on
the microphase morphology of a semicrystalline ABC triblock copolymer.
Macromolecules 1 999, 32, (12), 3994.

Kajiwara, A.; Maeda, K.; Kubo, N.; Kamachi, M., Origin of 16-line spectra
observed in ESR study of radical polymerization of tert-butyl methacrylate:
Clear evidence for ESR detection of polymeric propagating radicals.
Macromolecules 2003, 36, (3), 526.

49

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

L00, Y. L.; Register, R. A.; Ryan, A. J.; Dee, G. T., Polymer crystallization
confined in one, two, or three dimensions. Macromolecules 2001, 34, (26),
8968.

Ma, M.; Vijayan, K.; Hiltner, A.; Baer, E.; lm, J., Thickness effects in
microlayer composites of polycarbonate and poly(styrene-acrylonitrile).
Journal of Materials Science 1990, 25, (4), 2039.

Homminga, D. S.; Goderis, B.; Mathot, V. B. F.; Groeninckx, G.,
Crystallization behavior of polymer/montmorillonite nanocomposites. Part
III. Polyamide-6/montmorillonite nanocomposites, influence of matrix
molecular weight, and of montmorillonite type and concentration. Polymer
2006,47, (5), 1630.

Homminga, D.; Goderis, B.; Dolbnya, l.; Reynaers, H.; Groeninckx, G.,
Crystallization behavior of polymer/montmorillonite nanocomposites. Part
I. lntercalated poly(ethylene oxide)/montmorillonite nanocomposites.
Polymer 2005, 46, (25), 1 1359.

Homminga, D.; Goderis, B.; Dolbnya, I.; Groeninckx, G., Crystallization
behavior of polymer/montmorillonite nanocomposites. Part II. lntercalated
poly (epsiIon-caprolactone)/montmorillonite nanocomposites. Polymer
2006, 47, (5), 1620.

Valentini, L.; Biagiotti, J.; Kenny, J. M.; Santucci, 8., Effects of single-
walled carbon nanotubes on the crystallization behavior of polypropylene.
Journal of Applied Polymer Science 2003, 87, (4), 708.

Bansal, A.; Yang, H. C.; Li, C. 2.; Cho, K. W.; Benicewicz, B. C.; Kumar,
S. K.; Schadler, L. 8., Quantitative equivalence between polymer
nanocomposites and thin polymer films. Nature Materials 2005, 4, (9),
693.

Forrest, J. A.; Mattsson, J., Reductions of the glass transition temperature
in thin polymer films: Probing the length scale of cooperative dynamics.
Physical Review E 2000, 61, (1), R53.

Mattsson, J.; Forrest, J. A.; Borjesson, L., Quantifying glass transition
behavior in ultrathin free-standing polymer films. Physical Review E 2000,
62, (4), 5187.

Ferreiro, V.; Douglas, J. F.; Warren; Karim, A., Growth pulsations in
symmetric dendritic crystallization in thin polymer blend films. Physical
Review E 2002, 65, (5), 051606.

Ferreiro, V.; Douglas, J. F.; Warren, J. A.; Karim, A., Nonequilibrium

pattern formation in the crystallization of polymer blend films. Physical
Review E 2002, 65, (4), 042802.

50

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

Granasy, L.; Pusztai, T.; Warren, J. A.; Douglas, J. F.; Borzsonyi, T.;
Ferreiro, V., Growth of 'dizzy dendrites' in a random field of foreign
particles. Nature Materials 2003, 2, (2), 92.

ijillo, M.; Amal, M. L.; Muller, A. J.; Bredeau, S.; Bonduel, D.; Dubois,
P.; Hamley, l. W.; Castelletto, V., Thermal fractionation and isothermal
crystallization of polyethylene nanocomposites prepared by in situ
polymerization. Macromolecules 2008, 41, (6), 2087.

Maiti, P.; Okamoto, M., Crystallization controlled by silicate surfaces in
nylon 6-Clay nanocomposites. Macromolecular Materials and Engineering
2003, 288, (5), 440.

Prest, W. M.; Luca, D. J., Formation of Gamma-Phase from Alpha-
Polymorphs and Beta-Polymorphs of Polyvinylidene Fluoride. Journal of
Applied Physics 1978, 49, (10), 5042.

Wang, J. J.; Li, H. H.; Liu, J. C.; Duan, Y. X.; Jiang, S. 0.; Yan, S. K., On
the alpha -> beta transition of carbon-coated highly oriented PVDF
ultrathin film induced by melt recrystallization. Journal of the American
Chemical Society 2003, 125, (6), 1496.

Gersappe, 0., Molecular mechanisms of failure in polymer
nanocomposites. Physical Review Letters 2002, 89, (5), 058301.

Fekete, E.; Molnar, S.; Kim, G. M.; Michler, G. H.; Pukanszky, B.,
Aggregation, fracture initiation, and strength of PP/CaCO3 composites.
Journal of Macromolecular Science-Physics 1999, B38, (5-6), 885.

Si, M.; Araki, T.; Ade, H.; Kilcoyne, A. L. D.; Fisher, R.; Sokolov, J. C.;
Rafailovich, M. H.; 39, Macromolecules 2006, 39,4793.

Ray, S. S.; Yamada, K.; Okamoto, M.; Ueda, K., Control of
biodegradability of polylactide via nanocomposite technology.
Macromolecular Materials and Engineering 2003, 288, (3), 203.

Grady, B. P.; Pompeo, F.; Shambaugh, R. L.; Resasco, D. E., Nucleation
of polypropylene crystallization by single-walled carbon nanotubes.
Journal of Physical Chemistry B 2002, 106, (23), 5852.

Lei, M. 0.; Li, J.; Yang, J.; Liu, J. J.; Tong, X.; Cheng, H. M., The
morphology and thermal properties of multi-walled carbon nanotube and
poly(hydroxybutyrate-co-hydroxyvalerate) composite. Polymer
International 2004, 53, (1 0), 1479.

Probst, 0.; Moore, E. M.; Resasco, D. E.; Grady, B. P., Nucleation of
polyvinyl alcohol crystallization by single-walled carbon nanotubes.
Polymer 2004, 45, (13), 4437.

51

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.
93.

94.

95.

Anand, K. A.; Aganrval, U. S.; Joseph, R., Carbon nanotubes induced
crystallization of poly(ethylene terephthalate). Polymer 2006, 47, (11),
3976.

Kodjie, S. L.; Li, L. Y.; Li, B.; Cai, W. W.; Li, C. Y.; Keating, M.,
Morphology and crystallization behavior of HDPE/CNT nanocomposite.
Journal of Macromolecular Science Part B-Physics 2006, 45, (2), 231.

Phang, I. Y.; Ma, J. H.; Shen, L.; Liu, T. X.; Zhang, W. D., Crystallization
and melting behavior of multi-walled carbon nanotube-reinforced nylon-6
composites. Polymer International 2006, 55, (1), 71.

Wang, 8.; Sun, G. F.; Liu, J. J.; He, X. F.; Li, J., Crystallization behavior of
carbon nanotubes-filled polyamide 1010. Journal of Applied Polymer
Science 2006, 100, (5), 3794.

Ray, S. S.; Bousmina, M., Biodegradable polymers and their layered
silicate nano composites: In greening the 21st century materials world.
Progress in Materials Science 2005, 50, (8), 962.

lkada, Y.; Tsuji, H., Biodegradable polyesters for medical and ecological
applications. Macromolecular Rapid Communications 2000, 21, (3), 117.

Dorgan, J. R. L., H.J.; Palade, L.-l.; Cicero, Polylactides: properties and
prospects of an environmentally benign plastic from renewable resources.
Macromol. Symp. 2001, 175, 55.

Tokiwa, Y.; Calabia, B. P., Degradation of microbial polyesters.
Biotechnology Letters 2004, 26, (15), 1181.

Lenz, R. W.; Marchessault, R. H., Bacterial polyesters: Biosynthesis,
biodegradable plastics and biotechnology. Biomacromolecules 2005, 6,

(1). 1.
Doi, Y., Microbial Polyesters. 1st Ed .; VCH Publishers: New York, 1990.

Anderson, A. J.; Dawes, E. A., Occurrence, metabolism, metabolic role,
and industrial uses of bacterial polyhydroxyalkanoates. Microbiological
Reviews 1990, 54, (4), 450.

Lundgren, D. G.; Alper, R.; Schnaitm.C; Marchess.Rh, Characterization of
poly-Beta-hydroxybutyrate extracted from different bacteria. Journal of
Bacteriology 1965, 89, (1), 245.

Chen, G., G-Q., Polyhydroxyalkanoates, in Biodegradable polymers for

industrial applications, 1st Ed.; Woodhead Publishing Limited: Cambridge,
2005.

52

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.

Satkowski, M. M.; Melik, D. H.; Autran, J.-P.; Green, P. R.; Noda, I.;
Schechtman, L. A., Physical and processing properties of
polyhydroxyalkanoate copolymers. In Biopolymers, Volume 3B ,
Polyesters ll - Properties and Chemical Synthesis, Steinbiichel, A., Doi,
Y., Ed. Weinheim ; Chichester : Wiley-VCH: 2001.

Gordeyev, S. A.; Nekrasov, Y. P., Processing and mechanical properties
of oriented poly(beta-hydroxybutyrate) fibers. Journal of Materials Science
Letters 1999, 18, (20), 1691.

Gordeyev, S. A.; Nekrasov, Y. P.; Shilton, S. J., Processing of gel-spun
poly(beta-hydroxybutyrate) fibers. Journal of Applied Polymer Science
2001, 81, (9), 2260.

Briassoulis, D., An overview on the mechanical behaviour of
biodegradable agricultural films. Journal of Polymers and the Environment
2004, 12, (2), 65.

Bin, F.; Cheng, 0.; Hang, W.; Peng, S. W.; Wang, X. Y.; Dong, L. S.; Xin,
J. H., Modified poly(3-hydroxybutyrate-co-3-hydroxyvalerate) using
hydrogen bonding monomers. Polymer 2004, 45, (18), 6275.

Sato, H.; Paderrnshoke, A.; Nakamura, M.; Murakami, R.; Hirose, F.;
Senda, K.; Terauchi, H.; Ekgasit, 8.; Noda, l.; Ozaki, Y., Infrared
spectroscopy and X-ray diffraction studies on the structure and thermal
behavior of biodegradable polyhydroxyalkanoates. Macromolecular
Symposia 2005, 220, 123.

Saad, G. R., Blends of bacterial poly[(R)-3-hydroxybutyrate] with
oligo[(R,S)-3-hydroxybutyratel-diol. Polymer International 2002, 51, (4),
338.

Chan, C. H.; Kummerlowe, C.; Kammer, H. W., Crystallization and melting
behavior of poly (3-hydroxybutyrate)-based blends. Macromolecular
Chemistry and Physics 2004, 205, (5), 664.

Avella, M.; Rota, G. L.; Martuscelli, E.; Raimo, M.; Sadocco, P.; Elegir, G.;
Riva, R., Poly(3-hydroxybutyrate-co-3-hydroxyvalerate) and wheat straw
fibre composites: thermal, mechanical properties and biodegradation
behaviour. Journal of Materials Science 2000, 35, (4), 829.

Abou-Aiad, T. H.; El-Sabee, M. Z.; Abd-El-Nour, K. N.; Saad, G. R.; El-
Sayed, E. S. A.; Gaafar, E. A., Miscibility and the specific interaction of
polyhydroxybutyrate blended with polyvinylacetate and poly(vinyl acetate-
co-vinyl alcohol) with some biological applications. Journal of Applied
Polymer Science 2002, 86, (9), 2363.

53

106.

107.

108.

109.

110.

111.

112.

113.

114.

115.

116.

Wang, X.; Peng, S.; L., 0., Effect of poly(vinyl acetate) (PVAc) on thermal
behavior and mechanical properties of poly(3-
hydroxybutyrate)/poly(propylene carbonate) (PHB/PPC) blends. Colloid &
Polymer Science 2005, 284, (2), 160

Yoshie, N.; Asaka, A.; Inoue, Y., Cocrystallization and phase segregation
in crystalline/crystalline polymer blends of bacterial copolyesters.
Macromolecules 2004, 37, (10), 3770.

Kunze, C.; Freier, T.; Kramer, S.; Schmitz, K. P., Anti-inflammatory
prodrugs as plasticizers for biodegradable implant materials based on
poly(3—hydroxybutyrate). Journal of Materials Science-Materials in
Medicine 2002, 13, (11), 1051.

Chen, G. X.; Hao, G. J.; Guo, T. Y.; Song, M. D.; Zhang, B. H.,
Crystallization kinetics of poly(3-hydroxybutyrate-co-3-
hydroxyvalerate)lclay nanocomposites. Journal of Applied Polymer
Science 2004, 93, (2), 655.

Dekoning, G. J. M.; Lemstra, P. J., Crystallization phenomena in bacterial
poly[(R)-3-hydroxybutyrate] .2. Embrittlement and rejuvenation. Polymer
1993, 34, (19), 4089.

Comiber.J; Marchess.Rh, Physical Properties of Poly-Beta-
Hydroxybutyrate .4. Conformational-Analysis and Crystalline-Structure.
Journal of Molecular Biology 1972, 71 , (3), 735.

Yokouchi, M.; Chatani, Y.; Tadokoro, H.; Teranish.K; Tani, H., Structural
Studies of Polyesters .5. Molecular and Crystal-Structures of Optically-
Active and Racemic Poly(Beta-Hydroxybutyrate). Polymer 1973, 14, (6),
267.

Calos, N. J.; Kennard, C. H. L., Crystallinity of Poly(Beta-Hydroxybutyrate)
Using the Rietveld Technique. Polymer 1994, 35, (21 ), 4595.

Gazzano, M.; Focarete, M. L.; Riekel, C.; Scandola, M., Bacterial poly(3-
hydroxybutyrate): An optical microscopy and microfocus X-ray diffraction
study. Biomacromolecules 2000, 1, (4), 604.

Tanaka, T.; Fujita, M.; Takeuchi, A.; Suzuki, Y.; Uesugi, K.; Doi, Y.; lwata,
T., Structure investigation of narrow banded spherulites in
polyhydroxyalkanoates by microbeam X-ray diffraction with synchrotron
radiation. Polymer 2005, 46, (15), 5673.

Dekoning, G. J. M.; Scheeren, A. H. C.; Lemstra, P. J.; Peeters, M.;
Reynaers, H., Crystallization phenomena in bacterial poly[(R)—3-
hydroxybutyrate] .3. Toughening via texture changes. Polymer 1994, 35,
(21), 4598.

54

117.

118.

119.

120.

121.

122.

123.

124.

125.

126.

127.

128.

Barham, P. J.; Keller, A.; Otun, E. L.; Holmes, P. A., Crystallization and
morphology of a bacterial thermoplastic -poly-3-hydroxybutyrate. Journal
of Materials Science 1984, 19, (9), 2781.

Barham, P. J., Nucleation behavior of poly-3-hydroxy-butyrate. Journal of
Materials Science 1984, 19, (12), 3826.

El-Hadi, A.; Schnabel, R.; Straube, E.; Muller, G.; Henning, 8., Correlation
between degree of crystallinity, morphology, glass temperature,
mechanical properties and biodegradation of poly (3-hydroxyalkanoate)
PHAs and their blends. Polymer Testing 2002, 21, (6), 665.

An, Y. X.; Dong, L. 8.; Mo, Z. 8.; Liu, T. X.; Feng, Z. L., Nonisothennal
crystallization kinetics of poly(beta-hydroxybutyrate). Journal of Polymer
Science Part B-Polymer Physics 1998, 36, (8), 1305.

He, Y.; Inoue, Y., Effect of alpha-cyclodextrin on the crystallization of
poly(3-hyd roxybutyrate). Journal of Polymer Science Part B-Polymer
Physics 2004, 42, (18), 3461.

Kai, W. H.; He, Y.; Inoue, Y., Fast crystallization of poly(3-
hydroxybutyrate) and poly(3-hydroxybutyrate-co-3-hydroxy-valerate) with
talc and boron nitride as nucleating agents. Polymer International 2005,
54, (5), 780.

Liu, W. J.; Yang, H. L.; Wang, 2.; Dong, L. 8.; Liu, J. J., Effect of
nucleating agents on the crystallization of poly(3-hydroxybutyrate-co-3-
hydroxyvalerate). Journal of Applied Polymer Science 2002, 86, (9), 2145.

Kai, W. H.; He, Y.; Asakawa, N.; Inoue, Y., Effect of lignin particles as a
nucleating agent on crystallization of poly(3-hydroxybutyrate). Journal of
Applied Polymer Science 2004, 94, (6), 2466.

Krishnamoorti, R.; Vaia, R. A.; Giannelis, E. P., Structure and dynamics of
polymer-layered silicate nanocomposites. Chemistry of Materials 1996, 8,
(8), 1728.

Park, H. M.; Liang, X. M.; Mohanty, A. K.; Misra, M.; Drzal, L. T., Effect of
compatibilizer on nanostructure of the biodegradable cellulose
acetate/organoclay nanocomposites. Macromolecules 2004, 37, (24),
9076.

Kolstad, J. J., Crystallization kinetics of poly(L-Iactide-co-meso-lactide).
Journal of Applied Polymer Science 1996, 62, (7), 1079.

Schmidt, 8. C.; Hillmyer, M. A., Polylactide stereocomplex crystallites as

nucleating agents for isotactic polylactide. Journal of Polymer Science
Part B-Polymer Physics 2001 , 39, (3), 300.

55

129.

130.

131.

132.

133.

134.

135.

Ray, 8. S.; Okamoto, M., New polylactide/layered silicate
nanocomposites, 6 - Melt rheology and foam processing. Macromolecular
Materials and Engineering 2003, 288, (12), 936.

Ray, 8. S.; Maiti, P.; Okamoto, M.; Yamada, K.; Ueda, K., New
polylactide/layered silicate nanocomposites. 1. Preparation,
characterization, and properties. Macromolecules 2002, 35, (8), 3104.

Nam, J. Y.; Ray, S. S.; Okamoto, M., Crystallization behavior and
morphology of biodegradable polylactide/layered silicate nanocomposite.
Macromolecules 2003, 36, (19), 71 26.

Ray, S. S.; Okamoto, M., Biodegradable polylactide and its
nanocomposites: Opening a new dimension for plastics and composites.
Macromolecular Rapid Communications 2003, 24, (14), 815.

Kalaitzidou K, F. H., Askeland P, Drzal L.T., The nucleating effect of
exfoliated graphite nanoplatelets and their influence on the crystal
structure and electrical conductivity of polypropylene nanocomposites.
Journal of Materials Science, published online October 25, 2007.

Ruoff, S. R.; Ding, W.; Piner, R.; Stankovich, S., Remarkable mechanics
of extremely thin graphite platelets. Polymen'c Materials: Science &
Engineering 2005, 92, 511.

Cho, D.; Lee, S.; Yang, G. M.; Fukushima, H.; Drzal, L. T., Dynamic
mechanical and thermal properties of phenylethynyl-terminated polyimide
composites reinforced with expanded graphite nanoplatelets.
Macromolecular Materials and Engineering 2005, 290, (3), 179.

56

CHAPTER 2

NUCLEATION OF POLY(HYDROXYBUTYRATE) BY EXFOLIATED

GRAPHITE NANOPLATELETS

Introduction

PHB is a largely investigated polymer derived from renewable resources
which has not yet found extended industrial applications, mainly due to

81.2.3.4,5 iS

brittleness and a low processability window. The brittleness of PH
known to stem from three factors: a glass transition temperature close to room
temperature, secondary crystallization occurring upon storage at room
temperature, and extremely low nucleation density. All these factors are
interrelated, and derive from the high purity of PHB and its stereochemical
regularity.3

As described in Chapter 1, boron nitride (BN) is a commonly used efficient
nucleating agent for PHB. The structure of graphite is similar to that of BN, and
xGnP has the potential to provide a similar nucleating effect on the polymer.
Graphite can be intercalated and exfoliated into very thin nanosized platelets6

with extremely high surface area and in-plane stiffness as high as 1060 GPa7 ,

which can not only nucleate PHB, but also impart multifunctionality to the

57

polymer and expand the range of its applications. This chapter describes the
nucleating effect of xGnP on PHB.8

Understanding the nucleating effect of xGnP, specifically the kinetic and
morphologic aspects of the nonisothermal crystallization from the melt of PHB
and PHB/xGnP systems, is essential for optimizing the processing conditions for
thermoplastic materials under dynamic conditions such as extrusion and injection
molding.9'1°'"'12 A kinetic analysis of the nonisothermal crystallization from the
melt was undertaken using models based on modifications of the Avrami
equation. Optical microscopy and atomic force microscopy (AFM) were used to
characterize the morphologies of PHB/xGnP systems and to offer further insight

on the crystallization process.13'“"15

AFM was used to study the surface
morphology and nanostructure of the PHB/xGnP system offering insights into
processes such as crystallization, crystal thickening and crystal deformation,
phenomena which were also observed in situ.16

The study described in this chapter constitutes work published in 2007 in

Journal of Applied Polymer Science.8

Experimental details

Materials
PHB of 900,000 g/mol number average molecular weight was obtained from

Metabolix (Cambridge, MA), and used without further purification. The exfoliated

graphite nanoplatelets having an average size of 1 pm (xGnP-1) was produced

58

in-house‘s'17

starting with an acid intercalated graphite 160-50A produced by
Graftech (Cleveland, OH). The xGnP has a surface area of ~100 m2/g and
consists of platelets approximately 10 nm in thickness and 1 pm in diameter.
The platelet basal plane is a graphene sheet, uniform in structure and surface

energy with evidence of some oxygen functionalities located at the edges of

these basal surfaces.

Sample preparation
Samples of pure PHB and PHB containing, 0.01 % wt, 1 % wt, and 3 % wt

xGnP-1 were prepared by extrusion and injection molding. The PHB powder,
previously dried in a vacuum oven at 80 °C for three hours to remove the
moisture, was mixed with the xGnP-1 and extruded using a DSM micro-extruder
(15 cm3 capacity) (DSM Research B.V., The Nederlands). The temperatures in
the three heating zones of the micro-extruder were 175 °C, 185 °C, and 195 °C
respectively. The mixtures were extruded at 100 rpm, and the cycle time was
three minutes. The extruded polymer systems were transferred using a transfer
cylinder (kept at 190 °C) to a mini-injection molder and injection molded at a fixed

temperature of 50 °C.

Methods

3) Thermal analysis: differential scanning calorimetry

The melting and crystallization behavior of PHB/xGnP-1 samples were
investigated using a 2920 Modulated DSC (TA Instruments). The runs were

performed under nitrogen flow, after calibrating the DSC instrument using an

59

Indium standard. For each run, a fresh sample (5-10 mg) of polymer was heated
from room temperature to 190 °C, kept isothermal for five minutes, in order to
remove the thermal history, then cooled to —60 °C, reheated to 190 °C, and
cooled to room temperature. The heating/cooling rate was 10 °C/min for the first
heating/cooling cycle. Six different heating/cooling rates were used for the
second heating/cooling cycle: 5, 8, 10, 14, 17, and 20 °C/minute. The analysis
employed the data recorded during the second cooling step. During the
nonisothermal crystallization of PHB study, the samples were run consecutively,

using the same sealed aluminum pan as control to reduce error.

b) Optical microscopy

The samples for optical microscopy were prepared as thin polymer films by
heating very thin sections of PHB/xGnP-1 injection molded bars in a Mettler
Toledo FP82 hot stage. Specifically, the polymer was placed on glass or mica
slides, covered with glass or mica cover slips, and heated at a rate of 20
°C/minute from room temperature to 190 °C, kept isothermal for three minutes,
then either allowed to crystallize upon cooling to room temperature, or to
crystallize isotherrnally at different preset crystallization temperatures. The
morphology of PHB and the size of the spherulites were observed using an

Olympus BH2 optical microscope equipped with a SpotTM camera.

60

c) Atomic force microscopy

Thin polymer films were prepared as described for optical microscopy, using
the Mettler Toledo FP80 hot stage. All images were collected on a NanoscopeTM
lV instrument (Veeco, Santa Barbara, CA) equipped with a J scanner. Images
were collected in either contact or tapping mode, using commercially available
silicon or silicon nitride cantilevers. The AFM was initially calibrated by the
manufacturer but because of the nonlinear electrical response of piezo materials,
the electrical signals applied to SPM scanners were periodically calibrated to
ensure accuracy. Calibration standards, eg. the TGZO1 standard (MikroMasch,
CA) were imaged to ensure that the pitch and z distances are in agreement with
those specified by the manufacturer. Images were typically collected at a scan
rate of 0.5 Hz for scans sizes larger than 5XS um and up to 1 Hz for scan sizes
less than 1x1 pm. Some of the images were first-order flattened and scan lines

were erased.

d) Wide angle X-ray diffraction

X-ray diffraction traces of extruded and injection molded PHB/xGnP samples
were measured at room temperature in the scattering angle range 5° 3 26 s 45°
by using a Rigaku 200B X-ray imaging plate diffractometer equipped with a
curved graphite monochromator. The scanning rate was 0.020 ols. Cu Ka
radiation (wavelength 1.5418 A) from a Rigaku rotating anode X-ray generator

was used as an incident X-ray source (50 W, 100 mA).

61

Results and discussion

Crystallization of PHB_i_ruhe presence of xGnP-1

The nonisothermal crystallization of pure PHB and PHB containing 1 % wt,
and 3 "/0 wt xGnP-1 from the melt was studied using differential scanning
calorimetry. The melting temperature (Tm), the onset and end of crystallization
temperatures (To and T.,), the temperature of the crystallization peak (Tp), the
enthalpies of fusion and the crystallization enthalpies (AH, and Alfie) were
recorded for heating/cooling at six different rates (5, 8, 10, 14, 17, and
20 °C/minute respectively). Figure 2.1 shows the nonisothermal crystallization
curves for the PHB/xGnP-1 systems described above. After the addition of
xGnP-1, the crystallization peaks shifted to higher temperatures, which was the

first indication of the nucleating effect of xGnP-1.

62

 

Heat flow (arbitrary units)

 

 

 

 

30 50 70 90 110 130 150

 

 

Temperature (°C)

 

Figure 2.1 Nonisotherrnal crystallization curves obtained for (a) pure PHB, (b)
PHB/1% xGnP-1 and (c) PHB/3% xGnP-1 at six cooling rates:
5°C/min; 8°C/min; 10°Clmin; 14°C/min; 17°C/min; and 20°C/min
respectively. (The arrows indicate increasing cooling rates.)

The crystallization peaks also became narrower, which suggested that PHB
spherulites formed in the presence of xGnP-1 were smaller than the spherulites
formed by the neat polymer. Figure 2.2,a and Table 2.1 summarize the values of
T,D recorded for PHB containing different concentrations of xGnP-1, at different

cooling rates. Figure 2.2,b and Table 2.1 show that for PHB/xGnP-1 systems,

the degrees of supercooling (defined as the difference between the melting

63

temperature and the temperature of the crystallization peak, AT = Tm — Tp) were

lower than that of pure PHB. The nucleating effect of graphite nanoplatelets was
more evident at lower concentrations. Both the shifts of the crystallization peaks
and the decreases of the degree of supercooling are evidence of the nucleating

effect of xGnP-1.

 

 

§

 

 

 

 

 

 

 

 

120
a b

130 - 100 ~ V.;/'J
E3 .
E; 110 J W A 80
u. 0
a b 60 ~
g 90 . '2, 4o - M
E 70 -
n— 20 —

50 I I l 0 r r I T

o 5 1o 15 20 25 o 5 1o 15 20 25
Cooling rate (°Clmin) Cooling rate (°C/rrin)

 

 

Figure 2.2 Graphs showing (a) the temperature of crystallization peaks and (b)
the degrees of supercooling versus cooling rate for pure PHB (I),
PHB/0.01% xGnP-1 (o), PHB/1% xGnP-1 (o), and PHB/3% xGnP-1
(A)-

64

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0000.0 0000.0- 0000.0. 0000.0- 0000.0 00.0 00.0 00.00 00.000 000.00 00.000 00.0: 00
0000.0 0000.0. 0000.0. 0000.0- 0000.0 00.0 00.0 00.00 00.000 000.00 00.00 00.0: 00
0000.0 0000.? 0000.0- 0000.0- 0000.0 00.0 00.0 00 00.000 000.00 00.000 00.0: 00
0000.0 0000.0- 0000.0. 0000.0- 0000.0 0.0 00.0 00 00.000 000.00 00.000 8.0: 00
0000.0 0000.0- 0000.0. 080.0- 0000.0 00.0 00.0 00.00 000 05.00 00.000 00.000 0
0000.0 0000.0- 0000.0- 0000.0- 0000.0 00.0 00.0 00.00 0.000 00.000 00.0: 00.000 0
3:; FE; 6L 8L 8L 8L EL .5058
.N 0.0-002 0 00-002 0 00 a... 00 E0 .0 Q0 00 0
70580.08}.
0000.0 0000.0. 0000.0- 0000.0- 0000.0 0.0 00.0 00.00 00.000 00 00.000 000 00
0000.0 0000.0- 0000.9 0000.0- 0000.0 00.0 00.0 00.00 000 00 00.00 000 00
0000.0 008.0- 0000.0- 0000.0- 0:00 0.0 00.0 00.00 00.000 00 00.000 000 00
0000.0 0000.0- 0000.0- 0000.0- 0000.0 00.0 00.0 00.00 0.000 00 00.000 000 00
E00 0000.0- 0000.9 0000.9 0000.0 00.0 00.0 00.00 0.000 8 00.000 000 0
0000.0 0000.0- 0000.0- 0000.0- 0000.0 00.0 00.0 00.00 0.000 000 00.0: 000 0
E 3:; 6L 8L 8L 6L 8L Egg
.0 0.0-000. N 00-002 0 NE a... 00 E0 .0 00 00 Q
0-0550102.
0000.0 0000.0- 0000.9 0000.0 0000.0 00.0 00.0 00.000 00.000 00.00 00.00 00.000 00
0000.0 0000.0- - 0000.0 0000.0 00.0 00.0 00.00 00.000 00.00 00.00 00.000 00
@0000 0000.0- - 0000.0 0000.0 00.0 00.0 00.00 00.000 00.00 00.00 00.000 00
0000.0 0000.0- - 0000.0 0000.0 00.0 00.0 00.00 00.000 00.00 00.00 00.000 00
0000.0 0000.0- - 0000.0 0000.0 00.0 00.0 00.00 00.000 00.00 00.00 00.000 0
0000.0 0000.0- - 0000.0 0000.0 00.0 00.0 00.00 00.000 00.00 00.00 00.000 0
.005 E5. 8L 6L EL 8.0 8.0 EEBL
0N AuN-vmo_ N $-50. : 0:0 02¢ .2 E.» .K 0.0 0 .0. Q
0:0
9.209 2020.00. _Em0>< voEpoE h_o ENE—05w 0N 03a...

65

Kinetic analysis of the nonisflermal ongtallization from melt

Based on the shifts of Tp, the narrowing of the crystallization peaks, and the
decrease of the degree of supercooling of PHB after the addition of xGnP-1, it
was concluded that maximum crystallinity of PHB crystallizes was achieved
faster in the presence of xGnP-1. In order to verify this, a kinetic analysis was
performed to compare the nonisothermal crystallization from the melt of the
various PHB/xGnP-1 systems. Most methods used for describing nonisothermal
crystallization are based on modifications of Avrami equation (Equation
21)’18,10,11,12,19.20.

”'22 which accurately describes isothermal crystallization of

polymers:
1—Xt =exp(—Zt"), (2.1)

where X, is the relative degree of crystallinity, Z is an overall rate constant

characterizing both nucleation and growth contributions, and n is a mechanism
constant whose value depends on the type of nucleation and growth process
parameters.11 A modification proposed by Jeziomy,22 and applied to PHB,9
successfully described the kinetics of nonisothermal crystallization of
PHB/xGnP-1 samples. The cooling rate (D), assumed to be constant, was the
factor used in correcting the rate constant, Z, as shown in Equation 2.2. The

relative degree of crystallinity, X t , and the absolute degree of crystallinity at each

crystallization temperature, X c , were determined using (Equations 2.3 and 2.4).

loch = — (2.2)

66

T dH
H—der
T dT
x0: 0 0, (2.3)
(1‘ WxGnP-1)' AHf

 

 

(2.4)

where wxan_1 is the mass fraction of xGnP-1 in the polymer system,

AH? =146 Jlg is the thermodynamic melting enthalpy of 100 °/o pure crystalline

PHB,23 and AHc is the crystallization enthalpy recorded by DSC.

The development of relative crystallinity with temperature for the PHB/xGnP-1
systems investigated is presented in Figure 2.3.

In each of the three cases, after adding xGnP-1 to PHB, complete
crystallization was achieved over a narrower temperature range (Figure 2.3), and

also in a shorter time Table 2.1.

67

 

 

0.8 —
0.6 —
$2
0.4 —

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 2.3 Development of relative crystallinity with temperature for (a) pure
PHB, (b) PHB/1% xGnP-1, (c) PHB/3% xGnP-1 at six cooling rates:
5°Clmin; 8°C/min; 10°Clmin; 14°C/min; 17°C/min; and 20°C/min,
respectively. (The arrows indicate increasing cooling rates.)

68

 

 

Q8-

0.6 -
i ,0-
o.4 ~ , /

 

 

 

300 400

 

 

 

300 400

 

 

0.6 . x/
0.4 0
0.2 - f

 

 

 

o 100 200 300 400
NS)

 

 

 

Figure 2.4 Development of relative crystallinity with time for (a) pure PHB, (b)
PHB/1% xGnP-1, (c) PHB/3% xGnP-1 at six cooling rates: 5°Clmin;
8°C/min; 10°Clmin; 14°C/min; 17°C/min; and 20°C/min respectively
(The arrows indicate increasing cooling rates.)

69

The S-shaped curves indicate that pure PHB crystallizes in two steps, and
this phenomenon, also observed for ethylene terephtalate-ethylene oxide
segmented copolymers.11 is explained in terms of an initial fast growth phase of
the crystals in an amorphous environment, followed by a slower growth phase as
the crystals begin to impinge on each other in the interlamellar region.

The kinetic parameters for the nonisothermal crystallization PHB/xGnP-1

samples were determined by plotting log[—- In(1 - X t )] versus logt (Figure 2.5).

 

log(— In(1 - X t ))

 

 

 

 

logt
Cooling rates (°Clminute)
5-8-10—14—17—20

Figure 2.5 Graphs showing the results of modified Avrami analysis for (a)
PHB, (b) PHB/1% xGnP-1, and (c) PHB/3% xGnP-1 at six cooling
rates

70

FOr pure PHB, and PHB containing 1 "/0 wt and 3 °/o wt xGnP-1, the curves
were linear at first and then deviated from linearity, due to secondary
crystallization.1o However, for systems containing 0.01 % wt xGnP-1, the plots
were not linear, possibly due to non—unifonn dispersion of xGnP-1 in PHB, a
result of poor mixing in the micro-extruder. Hence, these samples were not
further subjected to kinetic analysis. The values of the kinetic parameters n, 2,
and Zc for neat PHB, and for PHB containing 1 °/o wt and 3 °/o wt xGnP-1 were
obtained from the slopes and intercepts of the lines Figure 2.5, and are
summarizedin Table 2.1. According to the modified Avrami analysis, the rate of

crystallization of pure PHB increased with increasing cooling rate (Figure 2.6).

 

 

 

 

 

 

 

1.2 l
:0 0.8 ~
IE
5
it} 0.4 -+

0,0 . . . .

0 5 10 15 20 25
Cooling rate (°Clmin)

 

Figure 2.6 Nonisotherrnal crystallization rate parameters for PHB/xGn-1
systems, determined according to modified Avrami analysis

For PHB samples containing 1 °/o wt and 3 % wt xGnP-1, the kinetic
parameters of the crystallization process were higher than for pure PHB, but

were independent of the cooling rate. Our results furthermore show that the time

71

required to achieve half of the final crystallinity (t1,2), as well as the maximum

crystallization time, were higher for pure PHB than for PHB containing exfoliated

graphite nanoplatelets (Figure 2.7,a, b, and Table 2.1.

 

 

 

 

 

 

 

 

 

 

4 8
a b

A 3 - ’q? 6 —
.8 8
D C
.E 2 -— 4 _
.E, g;
E 1 ~ .15' 2 —

o I I ' ' 0 r r fi f

o 5 1o 15 20 25 o 5 1o 15 20 25
Cooling rate (°Clmin) Cooling rate (°Clmin)

 

 

Figure 2.7 Graphs showing (a) the half-time crystallization times and (b) the
maximum crystallization times for pure PHB (I), PHB/0.01% xGnP-
1 (o), PHB/1% xGnP-1 (o), and PHB/3% xGnP-1 (A), determined
according to modified Avrami analysis

The activation energies for the nonisothermal crystallization from melt were
determined using the Kissinger method,1°'24 (Equation 2.5) based on the peak

temperatures recorded by DSC for different cooling rates (Table 1):

M— E (2.5)

72

where R is the universal gas constant, AE is the activation energy, and the other

parameters are as defined previously. The activation energies were calculated

using the slopes of the lines obtained by plotting log[

—2— vs l (Figure 2.8, and
Tp Tp
Table 2.1).

For PHB, PHB/ 1 °/o wt xGnP-1, and PHB/ 3 °/o wt xGnP-1, the activation

energies were found to be —60.41 kJ/mol, —213.61 kJ/mol, and —101.86 kJ/mol

respectively (Figure 2.9). AE for the nonisothermal crystallization from melt of

PHB was in agreement with the value of ~64.6 kJ/mol reported by An et al.9

 

 

 

 

 

 

 

-2

“E

0'254 _

‘0’

2

-2.3 . .

0.0024 0.0026 0.0028 0.003
1/Tp(K'1)

 

 

Figure 2.8 Kissinger plots used for determining the activation energy for

nonisothermal crystallization of pure PHB (I), PHB/0.01% xGnP-1
(o), PHB/1% xGnP-1 (o), and PHB/3% xGnP-1 (A)

73

 

 

AE (kJ/mol)

% wt xGnP-1 in PHB

 

 

 

Figure 2.9 Activation energies for the nonisothermal crystallization from melt of
PHB/xGnP-1 systems

Morpholoqv of PHB/xGnP-1 composites

Optical microscopy was used to observe the distribution and growth of the
crystallites formed by pure PHB and PHB/xGnP-1 systems during nonisothermal
cooling from the melt, and their morphologies after crystallization. As shown in
Figure 2.10,a, pure PHB formed large, well-defined banded spherulitic structures
having an average radius of 70 pm.

The optical micrographs obtained for PHB/1 % wt xGnP-1 and PHB/3 % wt
xGnP-1 (not shown) showed agglomerations of numerous, small spherulites
which crystallized spontaneously from the melt. This result was not unexpected
since the high concentration of xGnP-1 added (1 and 3% wt) introduced many
heterogeneous nuclei which initiated nucleation of the PHB at 110-113 °C. In an
attempt to observe the crystallization of PHB in the presence of xGnP-1, PHB

samples containing 0.01 °/o wt xGnP-1 were prepared. The addition of a lower

74

concentration of xGnP-1 to PHB led to the formation of spherulites now large

enough to be observed with optical microscopy (Figure 2.10,b).

 

Figure 2.10 Optical micrographs showing spherulites of (a) pure PHB and (b)
PHB/0.01 wt % xGnP-1 formed during nonisothermal crystallization
from the melt.

AFM was employed to provide further details regarding the sizes of the
spherulites and topographical information. AFM images were collected for the
pure PHB, and PHB containing 0.01 % wt, 1 % wt and 3 % wt xGnP-1
(Figure 2.11). Figure 2.11,a shows an AFM height image of a PHB spherulite
formed upon nonisothermal crystallization of PHB from the melt, having an

approximate diameter of 90 pm, which is most likely an overestimation due to tip

75

convolution effects. Figure 2.11,b shows an AFM height image for a spherulite
formed from the PHB/0.01 % wt xGnP-1 system also having a diameter of about
90 pm. In the case of the PHB/xGnP-1 systems having higher concentrations of
graphite nanoplatelets, the crystallites formed appeared to be much smaller and
impinged into each other (Figure 2.11,b,c), so despite the accuracy of AFM, due
to the nature of the sample, it was difficult to determine the exact sizes of the

spherulites.

 

1000 nm 800 nm

500 nm 400 nm

Onm

"2 150 nm 200 nm

 

 

 

 

o 5 10 15 pm

 

Figure 2.11 AFM tapping mode height images (top view) showing spherulites of
(a) pure PHB and PHB containing, (b) 0.01 wt % xGnP-1, (c) 1 wt
% xGnP-1 and (d) 3 wt % xGnP-1

76

These AFM results indicate that the Iamellar thickness of pure PHB
spherulites crystallized from the melt ranges from 7 to 20 nm, as shown in the

cross-sectional analysis in Figure 2.12,a.

 

 

 

 

 

 

 

 

 

 

 

 

45 *
100 nm
0 I I Afi
50 nm V WW
0 nm —45 -
I I l l l
0 2.0 4.0 6.0 8.0
nm pm
2.00 -
': 5.0 nm
0 [\A [1 MA I) A/\ A A
2.5 nm VV V V VV V V
0.0 nm -2.00 _
l l l
0 100 200
nm

 

 

 

Figure 2.12 AFM tapping mode height images (top view) and accompanying
cross-sections showing the Iamellar thicknesses of spherulites of
(a) pure PHB and (b) PHB/1 % wt xGnP.

The Iamellar structure was also observable in the PHB/1 % wt xGnP-1 system
at small scan sizes (250 x 250 nm) (Figure 2.12,b), and the thickness was
measured to be approximately 5 nm. On the other hand, AFM images of PHB/3
% wt xGnP-1 samples revealed an apparent disruption of the Iamellar structure,
attributed to the wide distribution of xGnP-1 throughout the crystalline structure

(Figure 2.13).

77

 

 

 

 

 

Figure 2.13 AFM height image (top view) of PHB/3 % xGnP-1 showing the
apparent disruption of the Iamellar structure of the polymer

Overall, these results indicate that the Iamellar thickness decreased as the
amount of xGnP-1 increased, and was proportional to the size of the crystalline
structures formed.

These experimental results are consistent with theoretical values for Iamellar
thicknesses, calculated using Equation 2.6, which relates Iamellar thickness to

the equilibrium melting temperature of polymers“:

 

20'
Tm = T,9,[1— M1fine)

where Tm is the observed meltingpoint, 0e is the fold surface free energy, AH is
the heat of fusion, and l is the Iamellar thickness. Using this correlation and the
values reported for Ga and AH (38 i 6 x10_3 J/m2 and

1.85x108 J/m3respectively,4 the Iamellar thickness, was estimated to be in the

range of 17 to 24 nm. Since the addition of xGnP-1 greatly increases the

78

heterogeneous nucleation ability of PHB, and hence the fold surface energy of
the PHB/xGnP-1 composite is expected to differ from pure PHB, the theoretical
estimation was not applied for PHB/xGnP-1 systems.

I.25 and Murase and coworkers,26 who

AFM was also employed by Skyes et a
used this technique to investigate solution-grown PHB single crystals prepared
from bacterial PHB by alkali hydrolysis, and reported the Iamellar thickness to be
approximately 5 nm.

WAXD measurements lead to the hypothesis of epitaxial growth of PHB onto
xGnP-1 surfaces. Figure 2.13 shows WAXD results obtained for PHB
nanocomposites containing various amounts of xGnP-1. For increasing amounts
of graphite, the relative intensity of the 020 and 110 diffraction peaks of PHB is
inverted, corresponding to a preferential growth of the crystals on the xGnP basal
plane. The intensity of the amorphous peaks recorded around 20 = 22 °
decreases as the amount of xGnP-1 in PHB increases, which is in agreement

with previously reported DCS measurements of the crystallinity of PHB/xGnP-1

nanocomposites.

79

 

 

 

 

“ ll 4“ 5 % xGnP-1
ll 11 4 % xGnP-1

 

 

 

 

g; __ -

U)

C

.9

.E

o

.2

‘6

g 0 3 % xGnP-1

U i 2.5 % xGnP-1

 

 

, A T ll 2% xGnP-1

AAA 1.5 % xGnP-1
M 1 % XGflP-‘l
‘ 0.5 % xGnP-1

_“ it A A” 1' ‘ , 0.1%xGnP-1

Neat PHB
5 10 15 20 25 30 35

 

 

 

 

 

 

 

 

 

20 (degrees)

 

Figure 2.14 Summary of WAXD analysis of PHB/xGnP-1 nanocomposites

This thermal behavior is particularly desirable for dynamic processing
conditions such as extrusion and injection molding, which are largely utilized for
obtaining biocomposites. The effects of xGnP on the crystallization behavior and
morphology of PHB provides a foundation for additional investigations of the
PHB/xGnP systems with controllable thermal and mechanical properties, and

electrical conductivity.

80

Conclusions

This research has shown that exfoliated graphite nanoplatelets efficiently
nucleate PHB, leading to systems that crystallize from the melt faster and at
higher temperatures. Very small amounts of xGnP-1 (0.01 % wt) increased the
crystallization temperature of PHB by approximately 30 °C, leading to the
formation of smaller spherulites, as evidenced by optical microscopy and atomic
force microscopy. A modified form of Avrami equation accurately described the
nonisothermal crystallization from melt of PHB/xGnP-1 systems, and showed that
this process occurred faster for PHB/xGnP-1 systems than for pure PHB. The
pure polymer crystallized faster as the cooling rate increased, while the kinetic
parameter describing the nonisothermal crystallization from the melt of rates of
the PHB/xGnP-1 systems were almost independent of the cooling rate.
Additionally, nonisothermal crystallization from the melt was shown to be more
energetically favored for the systems containing xGnP than for pure PHB.
Maximum crystallinity was also achieved faster in PHB/xGnP-1 systems than in
pure PHB. Although, the kinetic analysis and activation energies indicate that 3
wt °/o xGnP in PHB is excessive for initiating nucleation, it should be noted that
the excess xGnP may act as a nano-reinforcement, thereby leading to enhanced
mechanical properties and possibly achieving electrical conductivity in the

bionanocomposites through percolation of the xGnP network.

81

References

10.

11.

Briassoulis, D., An overview on the mechanical behaviour of
biodegradable agricultural films. Journal of Polymers and the Environment
2004, 12, (2), 65.

Dekoning, G. J. M.; Lemstra, P. J., Crystallization phenomena in bacterial
poly[(R)-3-hydroxybutyrate] .2. Embrittlement and rejuvenation. Polymer
1993, 34, (19), 4089.

Dekoning, G. J. M.; Scheeren, A. H. C.; Lemstra, P. J.; Peeters, M.;
Reynaers, H., Crystallization phenomena in bacterial poly[(R)-3-
hydroxybutyrate] .3. Toughening via texture changes. Polymer 1994, 35,
(21),4598.

Barham, P. J.; Keller, A.; Otun, E. L.; Holmes, P. A., Crystallization and
morphology of a bacterial thermoplastic -po|y-3-hydroxybutyrate. Journal
of Materials Science 1984, 19, (9), 2781.

El-Hadi, A.; Schnabel, R.; Straube, E.; Muller, G.; Henning, 8., Correlation
between degree of crystallinity, morphology, glass temperature,
mechanical properties and biodegradation of poly (3-hydroxyalkanoate)
PHAs and their blends. Polymer Testing 2002, 21, (6), 665.

Fukushima, H. Graphite nanoreinforcements in polymer nanocomposites.
Michigan State University, East Lansing, 2003.

Ruoff, S. R.; Ding, W.; Piner, R.; Stankovich, S., Remarkable mechanics
of extremely thin graphite platelets. Polymeric Materials: Science &
Engineering 2005, 92, 511.

Miloaga, D. G.; Hosein, H. A. A.; Misra, M.; Drzal, L. T., Crystallization of
poly(3-hydroxybutyrate) by exfoliated graphite nanoplatelets. Journal of
Applied Polymer Science 2007, 106, (4), 2548.

An, Y. X.; Dong, L. 8.; Mo, Z. S.; Liu, T. X.; Feng, Z. L., Nonisotherrnal
crystallization kinetics of poly(beta-hydroxybutyrate). Journal of Polymer
Science Part B-Polymer Physics 1998, 36, (8), 1305.

Liu, T. X.; Mo, Z. S.; Wang, S. G.; Zhang, H. F., Nonisotherrnal melt and
cold crystallization kinetics of poly(aryl ether ether ketone ketone).
Polymer Engineering and Science 1997, 37, (3), 568.

Kong, X. H.; Yang, X. N.; Zhou, E.; Ma, D. Z., Nonisotherrnal
crystallization kinetics of ethylene terephthalate-ethylene oxide segmented

82

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

copolymers with two crystallizing segments. European Polymer Journal
2000, 36, (6), 1085.

Kim, J. Y.; Park, H. S.; Kim, S. H., Crystallization behavior and nucleation
effect of carbon nanotube reinforced poly(ethylene 2,6-naphthalate)
nanocomposites. Polym. Mat: Sci. Eng. 2005, 92, 478.

lchikawa, M.; Nakamura, K.; Yoshie, N.; Asakawa, M; Inoue, Y.; Doi, Y.,
Morphological study of bacterial poly(3-hydroxybutyrate-co-3-
hydroxypropionate). Macromolecular Chemistry and Physics 1996, 197,
(8), 2467.

Hobbs, J. K.; Binger, D. R.; Keller, A.; Barham, P. J., Spiralling optical
morphologies in spherulites of poly(hydroxybutyrate). Journal of Polymer
Science Part B-Po/ymer Physics 2000, 38, (12), 1575.

Gazzano, M; Focarete, M. L.; Riekel, C.; Scandola, M., Bacterial poly(3-
hydroxybutyrate): An optical microscopy and microfocus X-ray diffraction
study. Biomacromolecules 2000, 1, (4), 604.

Hobbs, J. K.; Winkel, A. K.; McMaster, T. J.; Humphris, A. D. L.; Baker, A.
A.; Blakely, S.; Aissaoui, M.; Miles, M. J., Some recent developments in
SPM of crystalline polymers. Macromolecular Symposia 2001, 167, 1.

Cho, D.; Lee, S.; Yang, G. M.; Fukushima, H.; Drzal, L. T., Dynamic
mechanical and thermal properties of phenylethynyI-terminated polyimide
composites reinforced with expanded graphite nanoplatelets.
Macromolecular Materials and Engineering 2005, 290, (3), 179.

Grady, B. P.; Pompeo, F.; Shambaugh, R. L.; Resasco, D. E., Nucleation
of polypropylene crystallization by single-walled carbon nanotubes.
Journal of Physical Chemistry B 2002, 106, (23), 5852.

Avrami, M., Kinetics of phase change I - General theory. Journal of
Chemical Physics 1939, 7, (12), 1103.

Avrami, M., Kinetics of phase change ll - Transformation-time relations for
random distribution of nuclei. Journal of Chemical Physics 1939, 8, 212.

Liu, T. X.; Mo, Z. S.; Zhang, H. F., Nonisotherrnal crystallization behavior
of a novel poly(aryl ether ketone): PEDEKmK. Journal of Applied Polymer
Science 1998, 67, (5), 815.

Jeziorny, A., Parameters characterizing kinetics of nonisothermal

crystallization of poly(ethylene-terephthalate) determined by DSC.
Polymer1978, 19, (10), 1142.

83

23.

24.

25.

26.

Saad, G. R., Blends of bacterial poly[(R)-3-hydroxybutyrate] with
oligo[(R,S)-3-hydroxybutyratel-diol. Polymer lntemational 2002, 51, (4),
338.

Kissinger, H. E., Variation of peak temperature with heating rate in
differential thermal analysis. Journal of Research of the National Bureau of
Standards 1956, 57, (4), 217.

Sykes, K. E.; McMaster, T. J.; Miles, M. J.; Barker, P. A.; Barham, P. J.;
Seebach, D.; Muller, H. M.; Lengweiler, U. D., Direct imaging of the
surfaces of poly(Beta) hydroxybutyrate and hydroxybutyrate oligomers by
atomic-force microscopy. Journal of Materials Science 1995, 30, (3), 623.

Murase, T.; lwata, T.; Doi, Y., Atomic force microscopy investigation of
poly[(R)-3-hydroxybutyrate] Iamellar single crystals: Relationship between
molecular weight and enzymatic degradation behavior. Macromolecular
Bioscience 2001, 1, (7), 275.

84

CHAPTER 3

MODIFICATION OF CRYSTALLINITY AND xGnP/PHB INTERACTION

WITH IONIC LIQUIDS IN PHB/xGnP NANOCOMPOSITES

This chapter describes the potential use of ionic liquids in PHB/xGnP
nanocomposites, with a dual role of plasticizing the polymer matrix, and possibly

controlling the nucleating effect of xGnP.

Introduction

Ionic liquids (ILs) are a new class of environmentally friendly compounds.
They are classified as salts which melt at temperatures around 100°C, and have
an appreciable liquid range. Some of these compounds have melting points
around room temperature and are termed Room Temperature Ionic Liquids
(RTIL). lLs are sometimes defined “organic” salts because their structures are
based on big Organic cations. The most common cations and anions are
represented3 in Table 3.1. The degree and type of substitution of these bulky
asymmetric cations determines the low melting points of these salts, due to a

reduction of Coulombic interactions and disruption in the ion-packing and

85

reduction of the lattice energy. The anions control the compounds’ interaction
with water, the coordinating ability, and the hydrophobicity.

Besides the remarkably low melting points, ILs have no measurable vapor
pressure, are not flammable, and not oxidizing. Most of the applications of lLs
are related to their good solvent abilities, as these salts were proved to be good
solvents for a large range of organic and inorganic compounds. lLs fit very well
into the frame of green chemistry and are part of the general effort to reduce the
emission of volatile organic compounds.

The brittleness of PHB, which has a glass transition temperature close to
room temperature, is commonly addressed by addition of plasticizers. ILs were
shown to plasticize semicrystalline polymers such as PVC28 and PMMA" 2, which
indicates their potential to act as plasticizers for PHB. Since the mechanical
properties of PHB depend to a large extent on the crystallinity, and since the
presence of xGnP has been shown to have a significant effect on the
crystallization temperature, kinetics and crystalline morphology, imidazolium-
based ionic liquids (lLs) were investigated for their potential to modify PHB
brittleness as well as the interaction between the xGnP and the PHB polymer.
The large organic cations in lLs have a tremendous potential to interact with the
pi electrons on graphitic surfaces (through cation-7r interactions), leading to the
possible use of lLs for surface modification of xGnP. The adsorption of lLs onto
xGnP surface was expected to alter the nucleating ability and to offer more

control over heterogeneous nucleation processes.

86

Table 3.1 Common cations and anions in ionic liquids

 

 

 

 

 

 

 

 

 

 

 

 

MOST COMMONLY USED CA TIONS:
/®\ @ R1 (13 ' R1 7 69
|
./“V"\.. ~ 20.. i
l a2 0. 0/ gm
R _ _
1 ,3-dialkyl-imidazolium N-alkyl Tetraalkylammonium Tetraalkyl phosphonium
pyridinium
EXAMPLES OF ANIONS:
Water insoluble : Water soluble
[PFSI- [CH30021'
[(CF3802)2N]‘ [BF4]- [CF3002]. [N03]-
[(CzF5)3PF3]' [CFasoal‘ Br', C". I-
[BR1R2R3R41’ [AlzCl7l'. lA|C|4l‘

 

 

 

 

 

Effects of two imidazolium phosphates on PHB

The potential plasticizing effect of two ionic liquids, 1-ethyl-3-
methylimidazolium hexafluorophosphate (lL1) and 1-butyl-3-methylimidazolium
hexafluorophosphate) (lL2), was evaluated. PHB samples containing IL1 and IL2
were compared with PHB containing triethyl citrate (TEC), a common plasticizer
for thermoplastic biobased polymers. The various plasticized PHB systems were
compared in terms of mechanical properties (flexural strength and modulus, and
impact strength), glass transition temperature, as well as in terms of

polymer-plasticizer miscibility.

87

Common methods4' 5 described for the study of polymer miscibility are based
on the FIorry-Huggins theory. According to this theory, the equilibrium melting
temperature of a crystalline polymer may be depressed by monomeric diluents
incorporated in the polymer. The statistical thermodynamic treatment
incorporated in the method takes into account the nature of the polymer and the
concentration of the diluent, and also the degree of polymer-diluent interaction.

Equation 3.1 describes the Florry-Huggins interaction parameter".

-11. _ 2

where V1 is the molar volume of the plasticizer, R is the universal gas constant,
and 61 and 52 are solubility parameters of the plasticizer and of the polymer,

respectively. The smaller the Z. the stronger the polymer-diluent interaction and
consequently the better the solvent/plasticizer. For example, the Z=0-55

designates the dividing line between poor solvents and nonsolvents. Poor

solvency usually means 0.31 to 0.55, while 1 smaller than 0.30 implies a good

solvent.

Two methods for determining the interaction parameter were considered: a
method based on DSC measurements to determine the depression of the melting
temperature of the plasticized samples, and a second method4 which enables the
measurement of this temperature by using optical microscopy. The method
based on DSC measurements was selected and used for the determination of

FIorry-Huggins interaction parameter for PHB/lL1, PHB/ILS, and PHB/T EC

88

systems. This method was described for mixtures of PHB with oligo[(R,S)-3-
hydroxybutyrate]-dial? and is commonly known as Hoffmann-Weeks analysis.

According to this method, X is derived using Equation 3.2:

 

1 1 sz [ 02] [1 1] 2
.— =— x In— +——— +1 0 , (3.2)
T; T; Ava1 [ m2 M2 m1 % 121

where T,,°,' and Tn‘; are the equilibrium melting points of the blend and of the
crystalline polymer, AH;J is the heat of fusion for the 100 % crystalline polymer,

V1 and V2 are the molar volumes of repeating units and m and 0 represent the

degrees of polymerization and the volume fractions for each component.

Experimental details

Materials

PHB (number average molecular weight of 900,000 g/mol) was obtained from
Metabolix, Cambridge, MA. The melting point of this PHB sample was 170°C.
The polymer was used without further purification and dried at 80°C for three
hours in a vacuum oven to ensure the removal of trapped moisture.

1-ethyI-3-methylimidazolium hexafluorophosphate (Figure 3.1a) and 1-butyl-
3—methylimidazolium hexafluorophosphate (Figure 3.1b) were purchased from
Sigma Aldrich (product numbers 46093, and 70956, respectively). IL1 is solid at
room temperature and melts at approximately 58 °C. lL2 is a viscous liquid at

room temperature.

89

 

 

 

 

Figure 3.1 Structures of a) IL1, and b) "2

Tryethyl citrate (TEC) was obtained from Morflex, lnc., Greensboro, NC

(product name Citroflex 2).

Sample preparation

Samples of pure PHB and PHB containing up to 20 % (weight) ionic liquids or
triethyl citrate were prepared by extrusion and injection molding. The PHB
powder, previously dried in a vacuum oven at 80 °C for three hours to remove
the moisture, was mixed with the ionic liquids by hand and extruded using a DSM
micro-extruder (15 cm3 capacity) (DSM Research B.V., The Netherlands). The
temperatures in the three heating zones of the micro-extruder were 195 °C,
185 °C, and 175 °C respectively. The mixtures were extmded at 100 rpm, and
the cycle time was three minutes. The extruded polymer systems were injection

molded at a fixed temperature of 50 °C.

90

Methods

3) Mechanical properties evaluation

The flexural strength and modulus of elasticity of the neat and plasticized
PHB samples were determined using a United Test System machine, according
to ASTM Standard D790. The following testing conditions were chosen for all
determinations: load cell of 100 lb, gage length 2 in, test speed: 0.05 in/min.

The notched Izod Impact Strength was measured using an apparatus from
Testing Machines lnc., according to ASTM Standard 0256, using a load of 1 ft lb.
The sample size was 2.5 x 0.5 x 0.125 (in). Each result represents the average

obtained after testing five specimens.

b) Thermal analysis: differential scanning calorimetry and thennomechanical
analysis (TMA)

The melting and crystallization behavior of PHB, PHB/IL and PHB/T EC
samples were investigated using a 2920 Modulated DSC (TA Instruments). The
runs were performed under nitrogen flow, after calibrating the DSC instrument
using an Indium standard. For each run, a fresh sample (5-10 mg) cut from
injection molded bars was heated from room temperature to 200 °C, kept
isothermal for five minutes, in order to remove the thermal history, then cooled to
-60 °C, reheated to 200 °C, and cooled to room temperature. The heating and
cooling rates were 10 °C/minute.

DSC was also employed for undergoing Hoffman-Weeks analysis. Polymer

samples (5-10 mg) were heated from room temperature to 190 °C, kept

91

isothermal for five minutes, then cooled to different crystallization temperatures

(Tc), and allowed to crystallize isothennally. The crystallized samples were

further re-melted and the new melting points (Tm) corresponding to each Tc were
recorded.

The glass transition temperatures were also evaluated using a 2940 TMA (TA
Instruments). Samples for TMA were cut from injection molded bars using a

diamond saw.

Results and discussion

Modification of prperties of PHB 0 addition of plasticizers

Addition of IL1, lL2, and TEC respectively to PHB leads to decreasing of the
flexural strength and modulus, as shown in Figures 3.2 and 3.3. The same
decreasing trends were observed for samples containing ionic liquids as for
samples containing TEC. At the same time, the notched Izod impact strength of
PHB samples did not change significantly after addition of up to 20 % TEC
(Figure 3.4), while addition of similar amounts of IL1 and lL2 lead to diminished
impact resistance.

The glass transition temperature of PHB/IL1 and PHB/IL2 systems follows the
same decreasing trend with increasing amount of potential plasticizer observed

for PHB/T EC systems, as shown in Figure 3.5.

92

 

   
 
 

 

I
l
.
___....... .._._._.. .. 4

 

 

 

 

 

 

,0 50 -O--PHB+IL1
g +PHB+IL2 ,
2
0., 40+ +PHB+TEC
5
8
g 30-
(I)
t! 20—
3
2
u. 10*
0 . . . . 1
0 5 10 15 20 25

Plasticizer content (°/o wt)

 

 

 

Figure 3.2 Flexural strength of PHB samples containing plasticizing agents

 

 

 

-O- -PHB + IL1
+PHB + IL2
+PHB + TEC

   

 

 

 

Flexural modulus (GPa)
N

 

 

o l l I I
0 5 10 15 20 25
Plasticizer content (% wt)

 

 

 

 

Figure 3.3 Flexural modulus of PHB samples containing plasticizing agents

93

 

 

 

 

 

 

 

 

E
3
(0
(n
2
>
9
8
Lu 5‘ -O--PHB+IL1 —l—PHB+IL2
0 +PHB+TEC
o 5 10 15 20 25

 

Plasticizer content (% wt)

 

 

Figure 3.4 Notched Izod impact strength of PHB samples containing
plasticizing agents

 

 

 

 

 

 

 

 

 

 

 

15
-O- - PHB + IL1
10 ' —I— PHB + TEC
5 _ + PHB + lL2
8
L 0
,3
l 5
_5 i a
-10 —
-15
Plasticizer content (% wt)

 

 

 

Figure 3.5 Glass transition temperatures of PHB samples containing various
plasticizing agents

94

Several causes were suspected to determine the decrease of impact
strength: solidification of IL1 when cooling the polymer melt: degradation of lLs
during the processing of the polymer, inadequate processing conditions,
reactions of the lLs with the polymer, lack of miscibility of the polymer with lLs,
and changes in the morphology of the polymer. Further investigation determined
that the lack of miscibility of the lLs with PHB could be the most significant issue.
Hence, a miscibility analysis was performed, through the evaluation of

corresponding Florry-Huggins interaction parameters.

Application of Flony-Hgggins theory to analysis of miscibility of PHB with two
imidazolium phosphates

DSC measurements were performed using TA Instruments equipment, and
the equilibrium melting temperature was determined for each of the following
systems: PHB/IL1, PHB/10 % lL2, PHB/20 % lL1, PHB/ 5% IL2, PHB/ 13.8 %
lL2, and PHB/20 °/o lL2, according to Hoffmann-Weeks method described above.
For comparison, the analysis was done for the PHB/T EC systems (10 % and 20
% TEC in PHB respectively).

The equilibrium melting temperatures of the PHB/plasticizer systems were

obtained as the intersections of the Tm vs Tc lines with the corresponding Tm = Tc

line (Figure 3.6, and Figure 3.7). The obtained equilibrium melting temperatures

are presented in Table 3.2.

95

Table 3.2

Equilibrium melting temperatures of PHB samples containing

various plasticizing agents

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Polymer System Tmeq [oC]
PHB 182.93
PHB + 10% IL1 167.91
PHB + 20% IL1 170.49
PHB + 5% IL2 162.19
PHB + 13.8% lL2 162.97
PHB + 20% lL2 162.70
PHB + 10% TEC 180.67
PHB + 20% TEC 178.39
200
180 —
119.. ‘ I I ‘-
02 x X x W
. O ‘ A
160 1 . i ' ‘
I x a . >:
140 . . . . . . . .
40 60 80 100 120 140 160 180 200 220
Tc [°Cl
I PHB —Tm=Tc x PHB + 20%IL2
o Pl-IB+ 5% IL2 I PHB+13.8%|I2 1» PHB+10% TEC

 

A PHB + 20% TEC

0 PHB +10% IL1

X PHB 1' 20% IL1

 

 

 

 

Figure 3.6 DSC data collected for the determination of equilibrium melting
temperatures of PHB containing various potential plasticizers

96

Based on these temperatures, the calculated Florry-Huggins interaction
parameters were: 5.54 for PHB+10% TEC, 2.92 for PHB+20% TEC, and much
higher for the polymer systems containing lLs: 35.11 for PHB+5% IL2, 31.68 for
PHB+13.8% IL2, and 15 for PHB+20% lL2 . These results indicate a low
miscibility of PHB with IL1 and lL2, which is in agreement with the decreasing

impact strength measured for the corresponding PHB/IL injection molded

 

 

 

 

 

 

 

 

samples (Figure 3.4).
200
180 -
.— ‘ I
0 I
L. _ ,,,,, .,¢ : _ - .
E ------ E 'I ‘ ..... ’J .....
l- , .
160 1 "'
' /
I
140 l l l l l l l l
40 60 80 100 120 140 160 180 200 220
To [°C]
— Linear (PHB) — Linear (PHB + 20%IL2)
—— Linear (PHB + 13.8%ll2) — Linear (PHB + 10% TEC)
— - Linear (PHB + 20% TEC) — Linear (PHB + 5% lL2)
—Linear(PHB + 10% TEC) - - 'Linear (PHB + 10% IL1)
- - 'Linear (PHB + 20% IL1) —Linear (T m=Tc)

 

 

 

 

Figure 3.7 Extrapolated DSC data for PHB containing lL1, lL2, and TEC

97

 

Conclusions

The ionic liquids investigated in this work as potential plasticizers for PHB,
1-ethyI-3-methylimidazolium hexafluorophosphate and 1-butyl-3-
methylimidazolium hexafluorophosphate, were found to lack good miscibility with
the polymer matrix and to lead to decreased impact properties as compared to

pure PHB, and to PHB containing TEC.

PHB/1-butyl-4-methylimidazolium acetate/xGnP nanocomposites

The results regarding imizadolium hexafluorophosphates and PHB indicated
that the compounds screened for a plasticizing effect did not satisfactorily
compare to a traditional plasticizer. The search turned to an important
characteristic required for a good plasticizer for PHB, which is the ability to form
hydrogen bonds with the polymer.

Typically, the anion component of the IL controls the reactivity with water, the
coordinating ability, and the hydrophobicity. Common anions that yield useful lLs
include hexafluorophosphate [PF6]', tetrafluoroborate, [BF4]', bis(trifyl)imide,
[NTf2]', and chloride, Cl'. High symmetry pseudo-spherical, non-coordinating
anions are commonly regarded as optimal for formation of lLs, the existence of
low melting lLs containing anions such as methylsulfate°, dicyanamide7, and
bis(trifluoromethylsulfonyl)imide ([NTf2]')°' 8, show that shape and ion-interaction

factors are far from clear-cut. The [PF6]' and ([NTf2]‘) anions produce

98

hydrophobic solvents due to their lack of hydrogen-bond accepting abilityg,
though not all lLs containing these anions are hydrophobic.

The cations can hydrogen bond to the anion. The following scheme was
proposed by Wasserchield° for lLs containing buthyl-methylimidazolium ([bmimj")

cations, as described in Equations 3.3 - 3.6:

bmiml+J+-~A' <=> [bmim]...A (3.3)

, IEbmim]..[A I‘
K = 3.4
eqm mim” A“ ( )

The cations can also hydrogen bond to the polymer:
bmim1+l+ PHB c» [bmim]...PHB (3.5)

K, _ [[bmim].-.[PHB]]

eqm _ bmimtflpHB] (3'6)

 

This IL-polymer interaction was described for methyl methacrylate, and it
occurs via hydrogen bond formation between the cation and the carbonyl oxygen
of the electron withdrawing group. This clearly indicates that there is a
competition between the anion and the solute for the proton.

The focus remained on imidazolium-containing ionic liquids since they are
better characterized, and more readily available with a large range of anions.
1-butyl-4-methylimidazolium acetate (IL3) was chosen as a potential plasticizer
for PHB because the anion does not contain halogens, and is considered

non-toxic.

99

The adsorption of IL3 onto xGnP-1 surfaced, and effects of IL3 on PHB were

investigated and are reported here.

Experimental details

Materials

PHB (number average molecular weight 900,000 glmol, melting point 170 °C)
was obtained from Metabolix, Cambridge, MA. The polymer was used without
further purification and dried at 80 °C for three hours in a vacuum oven to ensure
the removal of trapped moisture. The exfoliated graphite nanoplatelets having an

average size 0f 1 pm (xGnP-1) was produced in-house'o'11

starting from
GrathuardTM 160-50A acid intercalated graphite produced by Graftech
(Cleveland, OH).

1-butyl-4-methylimidazolium acetate was purchased from Sigma Aldrich

(product number 39952).

Sample preparation
a) Surface treatment of xGnP-1 with lL3

1 g of dried xGnP-1 was immersed in 5 ml lL3, sonicated for 10 minutes, and
allowed to age for 24 hours at room temperature. The excess lL3 was removed
by washing with acetone/water, and the xGnP-1 particles were filtered and dried
in a vacuum oven for three hours in order to evaporate the remaining solvent.

b) Injection molded PHB/IL3 samples

100

Samples of pure PHB and PHB containing 1 % wt lL3 were prepared by
extrusion and injection molding. The PHB powder, previously dried in a vacuum
oven at 80 °C for three hours to remove the moisture, was mixed with IL3 and
extruded using a DSM micro-extruder (15 cm3 capacity) (DSM Research B.V.,
The Nederlands). For the extrusion of pure PHB, the temperatures in the three
heating zones of the micro—extruder were 175 °C, 185 °C, and 195 °C
respectively. The mixtures were extruded at 100 rpm, and the cycle time was
three minutes. The temperature of the mold was 50 °C. Various extrusion
conditions were attempted for PHB/IL3 samples, as described in the Results and

discussion section below.

Methods

3) Thermal analysis: Differential Scanning Calorimetry

The melting and crystallization of PHB, and PHB/IL3 samples were analyzed
using differential scanning calorimetry, using the same equipment and procedure

as described in the preceding section.

b) Molecular weight determination using dilute solution viscometry

Dilute solution viscometry is a technique commonly used to determine the
molecular weight of polymers by calculating the intrinsic viscosity of a dilute
solution of polymer in an adequate solvent, based on the efflux time. The
specific viscosity of the polymer solution in chloroform was determined based on

the efflux time (Equation 3.7 )2

101

Usp = —. (3])

where t is the efflux time of polymer solution, and to is the efflux time of pure

solvent. The reduced viscosities were calculated using Equation 3.8.

p
— ’ 3 o 8
’7red —C ( )

where C is the concentration of polymer in solution, and 77er is given in in dl/g.

The intrinsic viscosity, [1]] ,is determined from the intercept of the line obtained
by plotting the reduced viscosity versus concentration, and by applying the Mark-
Houwnik-Sakurada (M-H-S) equation (Equation 3.9) the molecular weight can be

estimated.
in]=K-M$. (3.9)

The M-H-S parameters for PHB12 are as follows: K = 7.7-10'5; a = 0.82, and

[77] = 7.7 -10'5 -Mf,’,'°2.

The maximum concentrations (Cmax) 0f PHB solution in chloroform and
PHB+1%IL3 in chloroform were 1.00 g/dl respectively. The solvent efflux time
was 133 seconds for each case.

DSV measurements were performed using Ubelohde tubes types Cannon 0C
an Cannon 0. The temperature was maintained at 77 °F using a heater and a

water bath. To ensure the reproducibility and accuracy of the results, the solvent

102

efflux time was measured three times for each concentration, and averaged for
each set of measurements.
For a solution volume of 14 ml, the following dilution scheme was applied:
— introduce 14.0 mL solution (Cmax);
— remove 4.0 mL and replace with 4.0 mL solvent (SCmax/7 = 0.714 Cmax);

—- remove 4.0 mL and replace with 4.0 mL solvent (25Cmax/49 = 0.510 Cmax);

- remove 4.0 mL and replace with 40 mL solvent (1250maxl343 = 0.364
Cmax);
— remove 4.0mL and replace with 4.0 mL solvent (625Cmaxl2401 = 0.260

Cmax);

— remove 8.0 mL and replace with 8.0 solvent (0.4 * last value = 0.104 Cmax).

c) X-Ray Photoelectron Spectroscopy

X-Ray Photoelectron Spectroscopy (XPS) was used for the evaluation of
xGnP-1, and xGnP-1 surface treated IL3. A Physical Electronics Phi 5400 ESCA
system with a non-monochromatic magnesium Kor X-ray source was used for
analyzing samples at pressures between 10’9 and 10”8 Torr, with a pass energy
of 29.35 eV, and a take-off angle of 45°. 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.

103

d) Evaluation of mechanical properties
The flexural strength, modulus, and the notched Izod impact strength of PHB,
PHB containing 3 % wt xGnP-1, and 3 % wt xGnP-1 surface treated with IL3

were measured as described in the first section of the present chapter.

Results and discussion

Processing of PHB in the presence of amounts of lL3 as low as 1 wt % was
not possible under the extrusion condition previously optimized for PHB alone.
The viscosity of the PHB/IL3 melt dropped dramatically upon heating in the
mini-extruder. The first indicator for decreasing viscosity was the drop in the

force measured in the extruder (Table 3.3)

Table 3.3 Processing conditions and forces recorded by the DSM extruder for
PHB, and PHB containing IL3

 

 

 

 

 

Temperature [°C] Force
Sample Remarks
Zone 1 Zone 2 Zone 3 [N]
PHB 195 185 175 1200 Optimized conditions
PHB+1% lL3 195 185 175 225
PHB+1% IL3 185 175 165 220 Not able to

 

PHB+1% IL3 175 175 175 200 extrude/injection mold.

 

PHB+1% IL3 175 165 165 200 Very low viscosity

 

 

 

 

 

 

 

PHB+1% IL3 175 165 165 200

 

104

 

The low PHB viscosity in the presence of lL3 indicated possible degradation
of PHB. Typical DSC runs (Figure 3.8) indicated that the melting point of PHB
dramatically decreased in the presence of 1 wt °/o IL3, from 173 °C to 146 °C,

while the crystallization of PHB from the melt was also affected.

 

30

 

 

 

 

 

 

202 PHB 104°C
102
o o.
s,
g -20-
E -30-
8 0.
I -504 PHB/1%wtlL3
173°C
50 . r . . . .
-100 -50 0 50 100 150 200 250

 

 

Temperature (°C)

 

Figure 3.8 DSC traces for PHB and PHB containing IL3

The dilution schemes for PHB and PHB containing 1 wt % IL3 in chloroform

are shown in Tables 3.4 and 3.5.

105

Table 3.4

Dilution scheme and dilute solution viscometry results for PHB in

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

chloroform
Concentration [gldl] Time t [s] 775,, 77.6,,
Cmax 1 .0500 404.33 2.0401 1 .9430
0.714 Cm...x 0.7497 349.67 1.6291 2.1730
0.510 Cmax 0.5355 300 1.2556 2.3448
0.364 Cmam 0.3822 270 1.0301 2.6951
0.260 Cmax 0.2730 245 0.8421 3.0846
0.104 Cmax 0.1092 191.33 0.4386 4.0165
Table 3.5 Dilution scheme and dilute solution viscometry results for PHB
containing 1 wt % IL3 in chloroform
Concentration [g/d I] Time t [s] 775,, 77.9,,
Cmax 0.1500 135.67 0.0463 Cmax
0.714 Cmax 0.1071 134 0.0334 0.714 Cmax
0.510 Cmax 0.0765 132.67 0.0231 0.510 Cm...x
0.364 Cmax 0.0546 132 0.018 0.364 CW,Ix
0.260 Cmax 0.0390 132.33 0.0129 0.260 Cmax

 

 

 

The intrinsic viscosity of PHB, calculated based on the measurements

presented in Figure 3.9, was [n]=3.7405, and corresponded to a weight

average molecular weight of 519,000 glmol.

106

 

 

— e PHB -—Linear (PHB)!

 

y = -1.9958x + 3.7405
R2 = 0.8121

 
   

Reduced viscosity (dg/I)
OD

 

 

O I I I I
0.00 0.20 0.40 0.60 0.80 1.00 1.20

c (d/dl)

 

 

 

Figure 3.9 Determination of the intrinsic viscosity of processed PHB

The intrinsic viscosity of PHB containing 1 °/o IL3, calculated based on the
measurements presented in Figure 3.10, was [77] = 0.2413, and corresponded to
a weight average molecular weight of 18,340 glmol. Hence, the molecular
weight of PHB decreased from approximately 519,000 glmol to approximately

18,340 glmol after the addition of 1 % (wt) of lL3.

107

 

0.50

0.45 7 o PHB+1%[BM|M]COOH

3:: 7 -— Linear (PHB+1%[BMIM]COOH)

0.30
0.251
0.202
0.152
0.10-
0.052
0.00 . .
0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60

 

 

 

 

 

y = -0.1183x + 0.2413
R2 = 0.7044

 

Reduced viscosity (dg/l)

 

 

 

I If

 

 

c (d/dl)

 

Figure 3.10 Determination of the intrinsic viscosity of processed PHB containing
1 wt % IL3

__urface treatment of xGnP-1 withfl

The surface treatment of xGnP-1 with IL3 was designed so that one could
introduce in the PHB/xGnP system a small amount of ionic liquid and one could
control the localization of the IL at the PHB/xGnP-1 interface. It was hoped that
the degradation of PHB in the presence of IL would affect the polymer chains in
the very proximity of the xGnP-1, and the resulted lower molecular weight
polymer at the interface with the carbon nanoplatelets will lead to better
properties of the nanocomposites.

The results of the XPS evaluation of the surface treatment are presented in
Table 3.6, which shows the atomic concentrations of xGnP-1 surfaces, as well as
of xGnP-1 surface treated with IL3. These results indicate that after the surface

treatment, the amount of nitrogen on the surface of xGnP-1 increases. This is

108

attributable to the nitrogen atoms in the imidazolium ring. The presence of
zirconium, only detected on the untreated xGnP-1, is due to the final stage of
xGnP-1 preparation, during which the size of xGnP is reduced from ~15 pm to ~1
pm in a vibratory mill that uses zirconium oxide milling balls was employed. No
Zirconium was detected on the surface treated xGnP-1, which indicates that this
residue was completely removed during the washing steps incorporated in the

surface treatment process.

Table 3.6 Atomic concentration for exfoliated graphite nanoplatelets surface
treated with lL3 (XPS results)

 

 

 

Sample C1 s N1s O1s Other
xGnP-1 92.62 0.45 6.71 Zr3d 0.22
XGnP/IL3 91.89 0.84 7.27 -

 

 

 

 

 

 

 

Mechanical properties of PHB/xGnP-1/IJL3 nanocomposites

The flexural strength, the modulus and the notched izod impact strength of
PHB, PHB containing 3 % wt xGnP-1 and PHB containing 3 % et xGnP-1 surface
treated with IL3 were measured. The results, presented in Figure 3.11, 3.12, and
3.13, indicate that the adsorption of lL3 onto the surface of xGnP did not lead to
improved mechanical properties of the corresponding PHB/xGnP-1

nanocomposites.

109

 

 

Flexural strength (MPa)
0 8 8 8 8 8 8

70

 

37

53

 

PHB 4' 3% xGnP PHB 4* 3% xGnP IL3

 

Sample

 

Figure 3.11 Effect of surface treatment of xGnP-1 with lL3 on the flexural
strength of PHB/xGnP-1 nanocomposites

Figure 3.12

Figure 3.13

 

 

Modulus (GPa)

 

  

 

 

PHB + 3%xGnP PHB + 3%xGnP IL3

 

Sample

 

Effect of surface treatment of xGnP-1 with IL3 on the modulus of
PHB/xGnP-1 nanocomposites

 

 

 

 

 

 

 

 

E 2° 15
5 15 13
8
—° 10
>
9
0 5
C
NJ

0

PHB + 3% xGnP PHB + 3%xGnP IL3
Sample

Effects of surface treatment of xGnP-1 with lL3 on the notched Izod

impact strength of PHB/xGnP-1 nanocomposites

110

Conclusions

Processing of PHB nanocomposites using ionic liquids was not reported
before this study. A review of the literature identified ionic liquids as potential
additives for PHB/xGnP nanocomposites that through plasticization can modify
the properties of the biopolymer, and through surface modification can be used to
change the surface of the exfoliated graphite nanoplatelets and hence the
polymer/xGnP interaction. The three ionic liquids investigated did not lead to
beneficial results on the properties of PHB/xGnP nanocomposites. The two
imidazolium phosphates were shown to be insufficiently miscible with the PHB,
which in turn lead to reduced impact properties. The imidazolium acetate, also
explored as a surface treatment agent for the xGnP, was proven to degrade
PHB, and to detrimentally affect the mechanical properties of the corresponding
PHB/xGnP nanocomposites. A fundamental understanding of the relation
between the structure of ionic liquids and their effect on the polymer, together
with the study of all the factors governing the system behavior, may eventually
allow the determination of ionic liquids which could, at the same time, act as
plasticizers for thermoplastic PHB matrix, and as surface modification agent for

the xGnP.

111

References

10.

11.

Scott, M. P.; Rahman, M.; Brazel, C. 8., Application of ionic liquids as low-
volatility plasticizers for PMMA. European Polymer Journal 2003, 39, (10),
1947.

Scott, M. P.; Brazel, C. S.; Benton, M. G.; Mays, J. W.; Holbrey, J. 0.;
Rogers, R. D., Application of ionic liquids as plasticizers for poly(methyl
methacrylate). Chemical Communications 2002, (13), 1370.

Welz-Biennann, U. in Ionic liquids — new materials for new applications,
Ionic Liquids Workshop, Tuscaloosa, AL, March 23-24, 2004.

Saad, G. R., Blends of bacterial poly[(R)-3-hydroxybutyrate] with
oligo[(R,S)-3-hydroxybutyratej-diol. Polymer International 2002, 51, (4),
338.

Anagnostopoulos, C. E.; Coran, A. Y.; Gamrath, H. R., Polymer—diluent
interactions. l. A new micromethod for determining polyvinyl chloride-
diluent interactions. Journal of Applied Polymer Science 1906, 4, (11),
181.

Wasserscheid, P.; Welton, T., Ionic Liquids in Synthesis VCH-Wiley:
Weinheim 2002.

MacFarlane, D. R.; Golding, J.; Forsyth, 8.; Forsyth, M.; Deacon, G. B.,
Low viscosity ionic liquids based on organic salts of the dicyanamide
anion. Chemical Communications 2001 , (16), 1430.

Rogers, R. D.; Seddon, K. R., Ionic Liquids; Industrial Applications for
Green Chemistry. American Chemical Society: Washington, DC, 2001.

Wilkes, J. S.; Zaworotko, M. J., Air and water stable 1-ethyl-3-
methylimidazolium based ionic liquids. Journal of the Chemical Society-
Chemical Communications 1992, (1 3), 965.

Fukushima, H. Graphite nanoreinforcements in polymer nanocomposites.
Ph.D. Dissertation, Michigan State University, East Lansing, 2003.

Cho, D.; Lee, 8.; Yang, G. M.; Fukushima, H.; Drzal, L. T., Dynamic
mechanical and thermal properties of phenylethynyl-tenninated polyimide
composites reinforced with expanded graphite nanoplatelets.
Macromolecular Materials and Engineering 2005, 290, (3), 179.

112

12.

13.

Marchesschault, R. H.; Okamura, K.; Su, C. J., Physical Properties of
Poly(Beta-Hydr3xy Butyrate) .2. Conformational Aspects in Solution.
Macromolecules 1970, 3, (6), 735.

Pukanszky, B.; Moczo, J., Morphology and properties of particulate filled
polymers. Macromolecular Symposia 2004, 214, 115.

113

CHAPTER 4

POLYLACTIC ACIDI EXFOLIATED GRAPHITE NANOPLATELET
NANOCOMPOSITES

Introduction

The nucleating effect of exfoliated graphite nanoplatelets on PHB, described
in Chapters 2 and 3, motivated the investigation of the consequences of the
heterogeneous nucleation on the properties of resultant nanocomposites. This
study refers to composites obtained from another semicrystalline polymer derived
from renewable resources, PLA, and xGnP. The investigations focus on the
relations between processing methods, dispersion of xGnP, and properties of the
nanocomposites. The results underline the importance of optimized mixing of
xGnP with PLA, and the immediate consequences of dispersion of xGnP on the
mechanical properties of the composites, and provide the basis for the
investigation of the nanoscale structure of PLA/xGnP nanocomposites
(Chapter 6).

Previous results obtained in the Drzal group for PP/xGnP"3 composites
indicate a strong dependence of the properties on the processing methods and

crystallization conditions.4' 5

114

We report here on the properties of PLA/xGnP composites prepared using
three different processing methods: extrusion and injection molding, pre-mixing

and compression molding, and solvent mixing and compression molding.

Experimental details

Materials

Injection molding-grade polylactide, PLA 3001 D, was obtained from
NatureWorkls®LLC (Minneapolis, MN), and used without further purification.
Exfoliated graphite nanoplatelets having an average size of 1 pm (xGnP-1) were
produced in-house as described in Chapter 2. PLLA was purchased from
Boehinger-lngleheim Chemicals (Petersburg, VA). Two different molecular
weight PLLA samples were used in this study: RESOMER® L209$ (intrinsic
viscosity 3.0 dl/g for 0.1 % solution in chloroform at 25 °C), and RESOMER®
L21OS (intrinsic viscosity 3.8 dl/g for 0.1 % solution in chloroform at 25 °C).

Methylene chloride and 2-propanol were purchased from Sigma Aldrich.

Sample preparation

a) Extrusion and injection molding
Samples of pure PLA 31000 and PLA 31000 containing, 1 % wt — 10 % wt
xGnP-1 were prepared by extrusion and injection molding. The PLA pellets,

previously dried in a vacuum oven at 80 °C for three hours to remove moisture,

115

were mixed with the xGnP-1 and extruded using a DSM micro-extruder (15 cm3
capacity) (DSM Research B.V., The Nederlands). The temperature in each of
the three heating zones of the micro-extruder was 200 °C. The mixtures were
extruded at 150 rpm, and the cycle time was three minutes. A transfer cylinder
heated at 193 °C was used to transfer the extruded polymer systems into the
injection molding machine. The optimized injection molding pressure and the

mold temperature were 80 psi, and 25 °C, respectively.

b) Pre-mixing and compression molding

In an attempt to improve the dispersion of xGnP-1 in polylactides, samples of
neat PLA L21OS and PLA L21OS/xGnP-1 nanocomposites containing 1 % wt to 8
% wt xGnP-1 were prepared by pre-mixing the polymer and the xGnP-1 in
isopropanol, a non-solvent for PLA, but a good dispersion medium for xGnP-1,
as follows. For preparing 1 g PLA L21OS/xGnP-1 nanocomposite mixture,
appropriate amounts of xGnP-1 were dispersed in 100 ml isopropanol by
sonication. A VirSonic 100 Cell Disrupter from Virtis (Gardiner, NY) equipped
with a 0.5 inch probe was used for 10 minutes. Corresponding amounts of PLA
L21OS were dispersed in isopropanol using the same conditions as above. The
xGnP-1/isopropanol and PLA L21OS/isopropanol dispersions were then mixed
together and sonicated for an additional 10 minutes. The isopropanol was
removed by overnight evaporation in the hood, then the PLAIxGnP-1 mixtures

were dried in a vacuum oven (20 hours at 80 °C) prior to compression molding.

116

The dried PLAIxGnP-1 mixtures were compression molded at 200 °C for 2
min, under an applied pressure of 18,000 psi, using a Carver laboratory press
(Carver Inc., Wabash, IN). The platens of the press were pre-heated to 200 °C

prior to compression molding.

c) Solution mixing and compression molding

PLA L2098, PLA L210S, and their composites containing with xGnP-1 were
also prepared by solution mixing and compression molding. The xGnP-1 loading
varied from 0 wt % to 9 wt % for both PLA samples (L2098 and 2108). For each
xGnP-1 loading, 1 g of PLAIxGnP-1 mixture was prepared as follows: appropriate
amounts of xGnP-1 were dispersed in 20 mL methylene chloride by sonication (5
min, 10 W) using a VirSonic 100 Cell Disrupter from Virtis (Gardiner, NY).
Corresponding amounts of PLA were further dissolved in the xGnP-1i methylene
chloride dispersions, and the obtained PLAIxGnP-1 mixtures were precipitated
using 2-propanol. The resulted PLA/xGnP-1 materials were kept at room
temperature overnight, and then dried in a vacuum oven at 80 °C for 20 h to
ensure the complete evaporation of the organic solvents. The dried PLAIxGnP-1
mixtures were compression molded at 200 °C for 2 min, under an applied
pressure of 18,000 psi, using a Carver laboratory press (Carver Inc., Wabash,
IN). The platens of the press were pre-heated to 200 °C prior to compression

molding.

117

Methods

3) Flexural strength and modulus evaluation

The flexural strength and modulus of the neat PLA 31000 and PLA
31000/xGnP-1 injection molded samples were determined using a United Test
System machine, according to ASTM Standard 0790. The following testing
conditions were chosen for all determinations: load cell of 100 lb, gage length 2

in, test speed: 0.05 in/min.

b) Dynamic mechanical analysis

The storage moduli of PLA L2098, PLA L21OS, and the corresponding and
composites with xGnP-1 were evaluated using a 2980 Dynamic Mechanical
Analyzer (TA Instruments, New Castle, DE). The samples (approximately 5.50
mm x 16 mm x 0.10 mm) were cut from the compression molded films, and
tested in tension during heating from room temperature to 120 °C. The reported
values are for storage modulus at room temperature, and represent the average

of three measurements per xGnP-1 concentration in the PLA.

c) Nanoindentation

The microscale determination of Young’s modulus and hardness of PLA
L21OS and PLA L21OS/xGnP-1 composites were conducted using a MTS-
Nanoindentation System (MTS Nano Instruments, Oak Ridge, TN) equipped with

a Berkovich type diamond tip. An XP module, allowing a maximum indentation

118

depth of 5000 nm, was employed. The samples investigated were PLAI
composite films of 01310.1 mm thickness. 25 indents were, 25 mm apart were

set for each sample, with a testing depth of 2000 nm.

d) Differential scanning calorimetry

The melting and crystallization of PLA 31000, PLA L2098, PLA L210S, and
their composites with xGnP-1 were analyzed using a 2920 Modulated DSC (TA
Instruments, New Castle, DE). The runs were performed under nitrogen flow,
after calibrating the DSC instrument using an Indium standard. For each run, a
fresh sample (6-10 mg) of polymer/ composite was placed in a seal aluminum
pan and subjected to two heating/ cooling cycles. During the first cycle,
incorporated in the DSC method to ensure the erasure of samples’ thermal
history, the samples were heated (10 °Clmin) from room temperature to 200 °C,
kept isothermal at 200 °C for 10 min, cooled (10 °Clmin) to 0 °C, and
equilibrated at 0 °C. The second heating/cooling cycle immediately followed and
involved heating (3 °Clmin) to 200 °C, keeping the samples isothermal at 200 °C,
then cooling (3 °C) to room temperature. Unless specified otherwise, the data

reported and analyzed refer to the second heating/ cooling cycle.

e) Electrical conductivity evaluation
Impedance Spectroscopy — two-point method - was employed to determine
the resistivity of PLAIxGnP-1 injection molded composites along the flow

direction. Samples (approximately 3x5x12 mm) were cut from the center parts of

119

injection molded bars using a diamond saw. In order to ensure good contact
between the polymer/composite samples and the electrodes, the composite
surfaces in contact with the electrodes were first subjected to oxygen plasma
treatment (ten minutes at 550 W), then coated with gold (1-2 nm coating
thickness). The electrical resistance was measured in the frequency range from
0.1 to 100,000 Hz. The reported results are average values for measurements
on five distinct samples for each xGnP-1 concentration in PLA, and represent
bulk electrical conductivity (converted from resistance by taking into account the

samples’ dimensions).

f) Optical microscopy

The samples for optical microscopy were prepared as thin polymer films by
heating very thin sections of PLNxGnP-1 injection molded bars in a Mettler
Toledo FP82 hot stage. Specifically, the polymer was placed on glass slide,
covered with glass cover slip, and heated at a rate of 20 °Clminute from room
temperature to 190 °C, kept isothermal for three minutes, then either allowed to
crystallize upon cooling to room temperature, or to crystallize isotherrnally at
different preset crystallization temperatures. The morphology of PLA and the
size of the spherulites were observed using an Olympus BH2 optical microscope

equipped with a Spot” camera.

120

Results and discussion

Properties of injection molded PLA 30001 D/xGnP-1 nanocomposites

Hot stage optical microscopy observations of PLA and PLA/xGnP-1 systems
crystallized from the melt under both isothermal and nonisothermal conditions
revealed that the injection molded grade PLA was nucleated by xGnP-1.
Figure 4.1 shows optical micrographs of PLA 3001 D and its composites with
xGnP-1 isotherrnally crystallized from the melt. The neat polymer did not show
any crystalline structures after being cooled from 200 °C and kept at 140 ° for
10 minutes. Under the same isothermal crystallization conditions, large
spherulites were observed in the presence of small amounts (0.01 % wt) of
xGnP-1 (Figure 4.1a). As the amount of xGnP-1 in the PLA increased,

spherulitic structures were no longer observable (Figure 4.1c,d).

 

Figure 4.1 Optical micrographs showing PLA/xGnP-1 composites crystallized
from the melt under isothermal conditions. The amount (wt %) of
xGnP-1 was a) 0 %; b) 0.01 % c)1 % and d) 10%

121

The crystallinity of extruded and injection molded PLA/xGnP-1 composites
containing 1 % wt xGnP-1 was 144 % higher than that of the neat PLA but
increasing amounts of xGnP-1 did not increase the crystallinity (Figure 4.2).
These results are similar to the results obtained for PHB/xGnP-1 systems

(Chapter 2).

 

Xcm (%)

 

 

 

xGnP-1 (% wt)

 

 

 

Figure 4.2 Degree of crystallinity of PLAIxGnP-1 composites prepared by
extrusion and injection molding

For PLA 3001D/xGnP-1 nanocomposites prepared by extrusion and injection
molding, an improvement in flexural modulus was only detected when the
concentration of xGnP-1 in PLA 3001D exceeded 4 wt % (Figure 4.3). The
maximum improvement in modulus was approximately 70 % for 8 - 9 % wt

xGnP-1 (~ 11 — 13 % vol). For comparison, 10 °/o vol xGnP-1 in a PP matrix lead

122

to approximately 200 % improvement in flexural modulus of the injection molded
PP/xGnP-1 nanocomposites.° The two systems are however difficult to analyze
side-by-side for several reasons. Both matrices are commercial-grade, and
might contain additives affecting their crystallization behavior. The polylactide
contains a mixture of D— and L- stereoisomers in an unknown ratio. The poly(D-
lactic acid) (PDLA) component increases the amorphous contents and limits the
ability of the matrix to crystallize. The PP, as a result of the polymerization
reaction, is likely to contain catalyst particles, which increase the nucleation
density of the neat polymer. If xGnP agglomerates at higher loadings and
polymer-rich regions are created in the composites, it is more likely for the PP
than for the PLA chains that are not in the vicinity of xGnP to nucleate and
crystallize. Nevertheless, the xGnP was shown to nucleate the ,B-form of PP
crystals“, which determines better mechanical properties than those of the more
common a- form, and account for further improvement in modulus.

Differences in dispersion of the xGnP-1 in PLA and PP might be due to the
melt flow index of the two polymers. The melt flow index at the processing
temperature is higher (up to 30 g/10 minutes) for the PLA used than for the PP
(12 gl10 minutes).

Micromechanical models used to predict the properties of polymer
composites7 severely overestimated both the value of PP/xGnP-1° and PLA
3001D/xGnP-1 composites.

The flexural strength of the PLA 3001D/xGnP-1 composites did not differ

significantly from that of neat PLA 30010 (Figure 4.3).

123

 

Modulus (GPa)

 

012345678910
xGnP—1(%wt)

 

 

 

Figure 4.3 Flexural moduli of PLA 30010 and PLA 3001D /xGnP-1
nanocomposites prepared by extrusion and injection molding

 

120 -
100 5
80 —
60 -
40 -
20 1

ol

Flexural strength (MPa)

 

012345678910
xGnP-1(%wt)

 

 

 

Figure 4.4 Flexural strength of PLA 30010 and PLA 3001D /xGnP-1
nanocomposites prepared by extrusion and injection molding

124

The percolation threshold for electrical conductivity in injection molded
PLA/xGnP-1 nanocomposites was approximately 9 % wt (~ 12 % vol)
(Figure 4.5). The electrical conductivity of PLA/xGnP-1 composites was found to
be similar to that of injection molded PP/xGnP-1 composites, which also had a

high percolation threshold of 9 % vol.

 

1.E-06
1.E-07-
1.E-085
1.E-09-
1.E-10—
1.E-11—
1.E-12-
1.E-13A
1.E-14
1.E-15 l l i l l l I . .

012345678910

xGnP-1(%wt)

 

Bulk conductivity (S/cm)

 

 

 

 

 

 

 

Figure 4.5 Bulk electrical conductivity of injection molded PLA 3001D /xGnP-1
nanocomposites

The exfoliated graphite nanoplatelets were shown to improve the modulus of
neat PLA, as well as to impart electrical conductivity, but at very high loadings.
The need of property improvement with smaller addition of nanoplatelets directed
the study towards exploring alternative processing methods, with emphasis on

dispersion of the nanoplatelets.

125

Properties of PLA L21OS/xgnP-1 namcorpmsites prepared by pre—mixing
and compression molding

The goal of pre-mixing the PLA with the graphite nanoplatelets was to
improve the dispersion of the xGnP.

The results of dynamic mechanical analysis of PLA L21OS and its composites
containing up to 11 % wt xGnP-1 did not indicate any improvement in the
modulus of PLA L21OS after the addition of xGnP, and even lower moduli for
samples containing over 4 % wt xGnP-1 (Figure 4.6). These poor mechanical
properties were attributed to agglomeration of the graphite nanoplatelets in the

PLA matrix.

 

 

O)

 

 

 

+ Storage modulus + Loss modulus

 

01
i

.5.
i
l
l

A I .
' I
l

Modulus (MPa)
N 00
l I
l l
l
l
l
l

_x
0 l
I
I
l
I

 

 

 

O
..
_
_
_.
_
_
._
-
_
_

4 5 6 7 8 9 10
xGnP-1(%wt)

O
.5
N
OD

 

 

 

Figure 4.6 Moduli of PLA L21OS and PLA L21OS/xGnP-1 composites prepared
by pre-mixing and compression molding, as determined by DMA

126

Nanoindentation experiments further confirmed the poor dispersion.
Figure 4.7 shows the moduli of PLA L21OS and its composites with xGnP-1, as
determined from nanoindentation. These results indicate a slight improvement of
modulus of the surfaces of PLA L21OS composites containing 4 % wt xGnP-1,
followed by a drop of moduli values for amounts of nanoplatelets exceeding 4 %
wt - a trend similar to the one obsenred for the DMA results. The decrease in
surface mechanical properties for higher amounts of xGnP—1 was observed for
the surface hardness measured concomitantly in the nanoindentation

experiments (Figure 4.8).

 

Modulus (GPa)
A

 

0 1 2 3 4 5 6 7 8
xGnP-1 (% wt)

 

 

 

Figure 4.7 Moduli of PLA L21OS and PLA L21OS/xGnP-1 composites, as
determined from nanoindentation experiments

127

 

0.45 7
0.40 7

0.30 7
0.25 7
0.20 7
0.15 7
0.10 7
0.05 7
0.00 7 0 'I
0 1 2 3 4 5 6 7 8

xGnP-1 (% wt)

0

(A)

01
l

Hardness (GPa)

 

 

 

 

Figure 4.8 Surface hardness of PLA L21OS and PLA L21OS/xGnP-1
composites, as determined from nanoindentation experiments

Both DMA and nanoindentation results for PLA L21OS/xGnP-1
nanocomposites prepared by pre-mixing and compression molding indicate poor
dispersion of xGnP in the polylactides. A DSC analysis of the nonisothermal
crystallization from the melt for PLA L21OS and composites containing xGnP-1
(Figure 4.9) showed that the neat PLA L21OS crystallizes from the melt at around
110 °C. Addition of 1 % xGnP-1 leads to a bimodal crystallization peak. The
same bimodal behavior was observed for higher amounts of xGnP-1 in the
polylactide, indicating two competing types of nucleating behavior — a
homogeneous nucleation mechanism, reflected by the peak occurring at lower
temperatures, and a heterogeneous nucleation mechanism, reflected by the

higher temperature nonisothermal crystallization peaks.

128

 

0.25

 

 

 

0%
1%

 

0.2 7

 

 

Heat Flow (W/g)

0.15 -

 

 

 

0.1 ‘Ii 7 .

70 90 1 10 1 30 1 50
Temperature (°C)

 

 

 

 

Figure 4.9 DSC traces for the nonisothermal crystallization from the melt of
PLA L21OS, and PLA L21OS containing xGnP-1

Figure 4.10 and 4.11 show that the two crystallization peaks are similar in
terms of peak area and fulI-width-at-half-maximum (FWHM) for 1 % wt xGnP-1 in
the polylactides. As the amounts of xGnP-1 increase, the higher-temperature
peak (corresponding to heterogeneous nucleation by the nanoplatelets),
becomes smaller than the accompanying peak (assigned to homogeneous
nucleation). This indicates amounts of xGnP-1 higher than 1 % wt lead to both
heterogeneous nucleation of the polylactides, and retardation of crystal growth.
Similar combination of nucleating effect and retardation of crystal growth was
observed in other nanocomposites obtained from semicrystalline polymers, for
example intercalated poly(ethylene oxide)/montmorillonite°, and

PCUmontmorillonite.9

129

 

 

  

 

 

-I—Area of homogeneous nucleation peak
+Area of heterogeneous nucleation peak
-0- Total area

Peak area
—|
-

 

 

 

 

 

 

o I I I I I I I I
0 1 2 3 4 5 6 7 8 9
% wt xGnP-1

 

 

 

Figure 4.10 Areas of peaks corresponding to nonisothermal crystallization from
the melt of PLA L21OS, and PLA L21OS/xGnP-1 composites

 

 

 

-I- FWHM oflwomogeneous nucleation peak

    

 

 

+ FWHM of heterogeneous nucleation peak

 

  

 

 

 

0 1 2 3 4 5 6 7 8 9
% wt xGnP-1

 

 

 

Figure 4.11 FulI-width-at-half-maximum values of the crystallization peaks
corresponding to PLA L210S, and PLA L21OS/xGnP-1 composites

130

The mechanical properties and the crystallization behavior of PLA
L210SIxGnP-1 composites prepared by pre-mixing and compression molding
indicated poor dispersion of the xGnP-1 in the polylactides, due to agglomeration
of the xGnP-1, and retardation of crystal growth. Figure 4.12 summarizes the
calculated degree of crystallinity of these composites, as calculated from using
the enthalpy of melting obtained from DSC scans. Addition 1 % wt xGnP-1 lead
to a 123 % increase in crystallinity as compared to the neat PLA. Amounts of
xGnP-1 higher than 1 % wt did not conduce to further increase in the overall
degree of crystallinity. The differences in mechanical properties are likely to be

related to the quality of the crystal structures formed.

 

 

Xcm (%)

 

 

 

 

0 1 2 3 4 5 6 7 8 9
xGnP-1(%wt)

 

 

 

Figure 4.12 Degree of crystallinity of PLAIxGnP-1 composites prepared by pre-
mixing and compression molding

131

Properties of P_LA_L_21OS/x§nP-1 nanocomsites prepared by solution
mixing andficomression moldi_ng

The use of solvents mediums to mix polymers and nanoparticles is a very
efficient way to achieve good dispersion of nanoparticles, but very frequently a
less green approach, mainly because of the toxicity of commonly used solvents,
and difficulty and cost of solvent recovery procedures.

For the preparation of PLA/xGnP composites, solution mixing was the last
method considered since the search for a solvent that at the same time dissolves
the polylactides and provides a good dispersion medium for the graphite
nanoplatelets lead to chlorinated compounds such as chloroform, and methylene
chloride, which are very volatile, and toxic.

Dynamic mechanical analysis revealed that the storage moduli of
PLA L21OS/xGnP-1 composites prepared by solution mixing and compression
molding was significantly higher than moduli of corresponding composites
prepared by pre-mixing and compression molding, as shown in Figure 4.13.

The analysis of the fracture surfaces of PLA/xGnP-1 composites containing
over 1 % wt xGnP-1 did not lead to observations that could be related to the
differences in the properties of the composites prepared using the two
approaches. Important differences between the pre-mixed and the
solution-mixed samples were revealed by differential scanning calorimetry, which
is a bulk characterization technique, and were further refined using a localized
thermal analysis method that was adapted for the characterization of PNC, as

described in Chapter 6.

132

 

 

 

 

7 4 I Solution mixing and compresion molding
6 Pre-mixing and compresion molding
To?
3 4 ~
3
E 3 1
O
8 2 2
.9.
(I) 1 _
o _
xGnP-1 (% wt)

 

 

Figure 4.13 Storage moduli of compression molded PLA L21OS and PLA
L21OS/xGnP-1 composites prepared using two mixing approaches

Conclusions

Solution mixing followed by compression molding leads to the best dispersion
of xGnP-1 in the PLA. Amounts of xGn_1 exceeding 3% wt lead to
agglomeration of xGnP and appear to determine two competing processes:
nucleation of the polymer matrix, and retardation of crystal growth. The
competition between the two processes, and the consequences on the
nanoscale structure of the PLA/xGnP-1 composites, are further described in

Chapter 6.

133

References

1. Kalaitzidou, K.; Fukushima, H.; Drzal, L. T., Multifunctional polypropylene
composites produced by incorporation of exfoliated graphite nanoplatelets.
Carbon 2007, 45, (7), 1446.

2. Kalaitzidou, K.; Fukushima, H.; Drzal, L. T., A new compounding method
for exfoliated graphite-polypropylene nanocomposites with enhanced
flexural properties and lower percolation threshold. Composites Science
and Technology 2007, 67, (10), 2045.

3. Kalaitzidou, K.; Fukushima, H.; Drzal, L. T., Mechanical properties and
morphological characterization of exfoliated graphite-polypropylene
nanocomposites. Composites Part a-Applied Science and Manufacturing
2007, 38, (7), 1675.

4. Kalaitzidou K, F. H., Askeland P, Drzal L.T., The nucleating effect of
exfoliated graphite nanoplatelets and their influence on the crystal
structure and electrical conductivity of polypropylene nanocomposites.
Journal of Materials Science 2007, published online October 25, 2007.

5. Miloaga, D. G.; Hosein, H. A. A.; Misra, M.; Drzal, L. T., Crystallization of
poly(3-hydroxybutyrate) by exfoliated graphite nanoplatelets. Journal of
Applied Polymer Science 2007, 106, (4), 2548.

6. Kalaitzidou, K.; Fukushima, H.; Miyagawa, H.; Drzal, L. T., Flexural and
tensile moduli of polypropylene nanocomposites and comparison of
experrnental data to Halpin-Tsai and Tandon-Wang models. Polymer
Engineering and Science 2007, 47, (11), 1796.

7. Hull, 0.; Clyne, T. W., An Introduction to Composite Materials. Second
Ed., Cambridge University Press, 1996.

8. Homminga, 0.; Goderis, B.; Dolbnya, I.; Reynaers, H.; Groeninckx, G.,
Crystallization behavior of polymer/montmorillonite nanocomposites. Part
I. lntercalated poly(ethylene oxide)lmontmorillonite nanocomposites.
Polymer 2005, 46, (25), 1 1359.

9. Homminga, 0.; Goderis, B.; Dolbnya, I.; Groeninckx, G., Crystallization
behavior of polymer/montmorillonite nanocomposites. Part II. lntercalated
poly (epsilon-caprolactone)/montmorillonite nanocomposites. Polymer
2006, 47, (5), 1620.

134

CHAPTER 5

THERMAL CONDUCTWITY OF POLYLACTIC ACID/EXFOLIATED

GRAPHITE NANOPLATELET NANOCOMPOSITES

Introduction

Incorporation of micro- and nano-size fillers with high thermal and electrical
conductivities in polymer matrices is expected to lead to composite materials with
improved thermal and electrical conductivity compared to the neat polymers.
This study describes the experimental determination of the thermal conductivity
of PLA composites with xGnP-1 and xGnP-15, and presents a new approach
suggested for the preparation of composites with high thermal conductivity using
morphology direction in biphasic polymer blends.

Carbon-based reinforcements are widely used for applications requiring
polymer composites with enhanced thermal and electrical performance. Carbon
fibers and carbon black are traditional fillers for polymer composites and were
shown to lead to materials with increased thermal conductivity. The thermal
conductivity along the graphene plane (Table 5.1 and Table 5.2) makes carbon-
based nanosize reinforcements, among which exfoliated graphite nanoplatelets,

attractive additives for PNC having high thermal conductivity. The most recently

135

reported results for room temperature thermal conductivity of a suspended
single-layer graphene1 indicate that a conductivity value in the range from

(4.8410.44)x1 03 to (5.30:0.48)x103 W/mK.

Table 5.1 Room temperature thermal conductivity of materials of various
crystalline structures

 

 

 

 

 

 

 

 

Material K (Wm'IK'1)
Diamond 600
Graphite (parallel to c-axis) 3550
Graphite (normal to c-axis) 89
SIC (single crystal) ~100
SIC (polycrystal) ~10-50
Epoxy resins 0.2 — 0.5
Polyesters 0.2 — 0.24

 

 

 

 

Nanosize fillers are largely used for improving thermal and electrical
properties of polymersz, the advantages of using smaller particles coming mainly
from the large surface area. Recently, Winey et al.3 summarized studies on
improvements in thermal conductivity of polymers by incorporating CNT,
underlining that while PNC incorporating CNT have low percolation thresholds for

electrical conductivity, the thermal conductivity fails to show similar improvement.

136

Table 5.2 Properties of various nanosize fillers used for improving thermal
and electrical conductivities of polymers

 

 

 

 

 

 

 

 

Smallest Electrical Thermal
. Shape . . . E . . . .
Flller dimensron Aspect ratio conductlwty conductrwty
(approx) (GPa)
(nm) (S/cm) (WImK)
Carbon
rod 50-100 50-200 500 700-1000 10-20
nanofiber
Carbon 100 — 500 —
rod 5—50 1,000 100 - 1000
MWNT 10,000 10,000
100 - 1000 -
Carbon SWNT rod 0.6 — 1.8 1,500 1000
10,000 10,000
Aluminisillicate
plate 1-10 50-100 200—250 1—10
nanoclay
Nano—Ti02 sphere 10 -40 ~ 1 230,000 10'11 - 10'12 12
Nano-Al203 sphere 50-100 ~ 1 50 20 - 30 20 - 30

 

 

 

 

 

 

 

Extensive research was devoted to understanding the reasons why in
polymer/CNT composites the thermal conductivity improvement is not as sharp
as the corresponding improvement in electrical conductivity. The main reason4
for the difference between thermal and electrical transport in composites
incorporating high aspect ratio CNT is thought to be the relatively small
difference between the thermal conductivity of the matrix and the thermal
conductivity of the CNT (kcm/lpolymer ~ 104). The corresponding ratio of electrical
conductivities is ~ 1012 - 101°. The heat transfer is more likely to involve the
polymer matrix. Secondary reasons are the interfacial thermal resistance for the

heat flow between the CNT and the polymer/carbon nanotube interfacial thermal

137

 

resistance. Huxtable et al.° determined that in polymer/CNT composites the
nanotubes-polymer interfacial thermal resistance is 12 MWm'zK'1, and the
interfacial thermal resistance between CNT is also very high.

Heat transfer through materials occurs through transmission of phonons —
lattice vibrations -, and free electrons (if present). The phonons and the free
electrons are heat carriers. From a kinetic point of view, these carriers have a

mean free path It between collisions (energy exchange events), and an average

velocity v. The kinetic theory defines the thermal conductivity, K (Equation 3.1).

K = %cwt, (5.1)
where c is the volume specific heat of the carrier concemed.° The carriers’
average velocities are very insensitive to temperature. Since polymers do not
have free electrons, in order to understand the mechanism of heat transfer in
polymer composites one needs to focus on the phonons as carriers. Phonon
velocity (speed of sound) is higher in light, stiff materials. The mean free path of
phonons is very structure-sensitive and can be very high in specimens of high
perfection and large grain size. Table 5.1 offers some examples in this respect.
Crystalline polymers have slightly higher thermal conductivity than amorphous
ones. For filled semicrystalline polymers, it is extremely important to understand
how the fillers, regardless of their size, affect the crystalline structure, since this
governs the transport properties.

PLA and PHB are thermoplastic polyesters and act as insulators for electricity

and heat. Their transport properties have a direct relationship to their crystalline

138

structure, and can be tailored by incorporating fillers specific to desired
applications. Exfoliated graphite nanoplatelets have a high thermal conductivity
and when combined with semicrystalline PLA matrix, nanocomposites with
enhanced thermal conductivity are expected. The present study aims to
compare the effect of exfoliated graphite nanoplatelets of two different sizes on
the thermal and electrical conductivities of PLA.

The experimental results obtained for the thermal conductivity of PLA/xGnP
composites presented here indicated the possibility of a percolating mechanism
for the thermal conductivity and the need for strategies that could allow control
over morphology and preferable alignment of graphite nanoplatelets along
continuous paths.

Several examples are reported in the literature where composites of
conductive fillers and nearly immiscible biphasic polymer blends (containing
usually a semicrystalline polymer and an amorphous polymer) showed very good
control of electrical conductivity and low percolation thresholds. In such biphasic
systems, the conductive filler localizes either in the semicrystalline matrix
(Figure 5.1a) or at the interface between the amorphous polymer and the

semicrystalline polymer (Figure 5.1b).7'9

139

 

Figure 5.1 Localization of carbon black particles (1 % wt) in the PP phase of a
PP/ PS blend (a), and at the interface between PP and PS in a
similar blend (b)

This chapter presents results obtained using this approach for achieving

morphology control in PLA/xGnP composites.

Experimental details

Materials

Injection molding grade polylactic acid, PLA 30010, was obtained from
Nature Works LLC (Minneapolis, MN) and used without purification. Poly(L-lactic
acid), RESOMER® L21OS, was purchased from Boehinger-lngleheim Chemicals
(Petersburg, VA). The intrinsic viscosity of this polymer sample is 3.8 dI/g for
0.1 °/o solution in chloroform at 25 °C). Exfoliated graphite nanoplatelets having
an average size of 1 pm (xGnP-1) was produced in-house from GrafGardTM

160_50A acid intercalated graphite produced by Graftech (Cleveland, OH), as

140

described in Chapter 2. Methylene chloride and 2-propanol were purchased from
Sigma Aldrich. Polystyrene with weight average molecular weight 280,000 glmol

was purchased from Sigma Aldrich.

Sample preparation
a) Preparation of PLA 3001D/xGnP-1 composites

PLA composites with exfoliated graphite nanoplatelets (PLA 3001D/xGnP-1)
were prepared by extrusion and injection molding, as described in the beginning

of this chapter.

b) Preparation of PLA L21OS/xGnP-1 composites

PLA composites with exfoliated graphite nanoplatelets (PLAIxGnP-15 and
PLAIxGnP-1) were prepared by solution mixing and compression molding, as
described in the beginning of this chapter. The xGnP loading varied from 0 wt %

to 30 wt %. The thickness of the composites was 0.13 - 0.20 mm.

c) Preparation of PLA 21OS/PS/xGnP-1 composites

Composites of xGnP-1 with a mixture of PLA L21OS and PEG were prepared
by solution mixing (in methylene chloride) and compression molding. xGnP-1
was dispersed in methylene chloride, PLA L21908 and PS were dissolved in
methylene chloride, then the xGnP-1 dispersion and the polymer solution were

mixed by ultrasonication as described before. The composite mixture obtained

141

after evaporation and drying overnight in a vacuum oven was compression

molded under the same conditions as the PLA L210S/xGnP-1 composites.

Methods

a) Guarded heat flow method for thermal conductivity determination

The thermal conductivity of PLA and its composites with xGnP-1, and
xGnP-15 was measured using a UNITHERMTM machine - UNITHERMTM Model
2022 (Anter Corporation, Pittsburgh, PA). The tests were performed according to
ASTM E1530 (Standard Test Method for Evaluating the Resistance to Thermal
Transmission of Materials by the Guarded Heat Flow Meter Method Technique).
Specimens of one inch diameter were cut from the compression molded samples
using a punch. In order to ensure that the sample thickness was within the
recommended range, 2 to 10 discs were stacked-up for the composites with
higher xGnP loading. The samples were tested at 30 °C under an applied load of
30 psi. Reported results represent the average of three measurements for each

xGnP loading.

b) Two-point method for evaluation of surface conductivity

Impedance Spectroscopy — two-point method — was employed to determine
the resistivity of PLA/xGnP, PLAIPSIxGnP, and PLA/PEG/xGnP composites
along the flow direction. Samples (approximately 0.13x5x12 mm) were cut from
the center parts of compression molded films. The electrical resistance was
measured in the frequency range from 0.1 to 100,000 Hz, as described in the first

section of the present chapter. The reported results are average values for

142

measurements on five distinct samples for each xGnP-1 concentration in PLA,
and represent bulk electrical conductivity (converted from resistance by taking

into account the samples’ dimensions).

0) Dynamic mechanical analysis (DMA)

The storage moduli of PLA L21OS, PS, and their composites with xGnP-1 and
xGnP-15 were evaluated using a 2980 Dynamic Mechanical Analyzer (TA
Instruments, New Castle, DE). The samples (approximately 5.50 mm x 16 mm x
0.10 mm) were cut from the compression molded films, and tested in tension
during heating from room temperature to 120 °C. The reported values are for
storage modulus at room temperature, and represent the average of three

measurements per xGnP-1 concentration in the PLA, or PLA/PS blends.

d) Hot stage optical microscopy

The samples for optical microscopy were prepared as thin polymer films by
heating very thin sections of PLAIPSIxGnP-1 composites in a Mettler Toledo
FP82 hot stage. Specifically, the polymer was placed on glass slide, covered
with glass cover slip, and heated at a rate of 20 °Clminute from room
temperature to 190 °C, kept isothermal for three minutes, then either allowed to
crystallize upon cooling to room temperature. The morphology of PLAIPSIxGnP
was observed using an Olympus BH2 optical microscope equipped with a Spot”

camera.

143

Results and discussion

The maximum improvement (44.8 %) in thermal conductivity of extrusion
mixed/injection molded PLA 3001D/xGnP-1 composites was obtained for 9 % wt
xGnP-1 in the polylactide matrix (Figure 5.2). For these composites, the
percolation threshold for electrical conductivity was also very high, at about

9 % wt (Figure 4.3).

 

0.3

 

 

 

 

 

0.1 - L . i . i L . i .
0 1 2 3 4 5 6 7 8 9 10
%wthnP-1

 

 

 

Figure 5.2 Room temperature thermal conductivity of extrusion (melt) mixed
injection molded PLA 3001D/xGnP-1 composites

Higher improvement in thermal conductivity was observed in the
PLA L21OS/xGnP-1 composites prepared by solution mixing and compression
molding. This preparation method also allowed incorporation of much higher

amounts of xGnP-1 in the polylactide matrix. For 30 % wt xGnP-1 (highest

144

loading), the improvement in thermal conductivity was as high as 300 %

(Figure 5.3).

 

1.6

1_4 _. _ .. _. . + PLA L210$lenP-1
-I- PLA L21 OS/xGnP-1 5

 

 

 

 

 

 

 

 

 

0 5 10 1 5 20 25 30
xGnP-1 (% wt)

 

 

 

Figure 5.3 Room temperature thermal conductivity of PLA L210S/xGnP
composites prepared by solution mixing and compression molding

Figure 5.3 also shows that the improvement in thermal conductivity obtained
for the 30 % wt xGnP-1 was achieved with approximately 15 % wt of the larger
xGnP-15. The larger diameter nanoplatelets lowered the thermal contact
resistance of the polymer/graphite interfacial surface to the total conductivity, as
explained in Chapter 1 (Equation 1.1).

Apart from the xGnP particle size determining the extent of the contribution of
the interfacial layer to the bulk thermal conductivity, it appeared that the PLA
L21OS/xGnP systems presented a percolating behavior for thermal conductivity,

similar to the percolation behavior for electrical conductivity (Figure 5.4).

145

 

 

1.E+01
1.E-O1 -
1.E-03 ‘
1.E-05 -

   

1.E-07 -

   

I

Conductivity (sin‘)

1.E-09 -

 

 

+ PLA L21OS/xGnP-1
-I- PLA L210SlenP-15
1.E-13 . . i . l 0.
0 5 10 15 20 25 30
xGnP (% wt)

 

 

 

 

 

 

 

 

Figure 5.4 Specific electrical conductivity of PLA L21OS/xGnP-1 composites
prepared by solution mixing and compression molding

Figure 5.4 shows the specific electrical conductivities of the same PLA L21OS
composites with xGnP-1 of two different sizes, and illustrates the similarity
between the trends in thermal and electrical conductivities. These similarities,
together with the results obtained for PLA 3001D/xGnP-1 composites, indicate a
possible percolating mechanism for thermal conductivity, relying on the
establishment of continuous xGnP network.

Both the electrical and the thermal conductivities of the composites improved
when the amount and dispersion of xGnP in the matrix allowed contact between
neighboring platelets. At high xGnP loadings, percolated networks in the
semicrystalline PLA matrix could be created by the graphite nanoplatelets that

were in excess for nucleation purposes and remained outside the newly formed

146

crystallites. The percolation threshold for electrical conductivity was lower for the
larger platelets, which were more likely to touch each other in the polymer matrix.

As expected, for the same amount of xGnP in the PLA matrix, the
improvement in thermal conductivity was much smaller than the improvement in
electrical conductivity.

We have previously shown10 (Chapter 2) that amounts of xGnP-1 higher than
3 % wt determine dramatic decrease in the size of the polymer crystallites, and
the polymer crystals nucleated by the graphite nanoplatelets are more irregular.
Unlike during electric transport — when electrons travel through the percolating
tube network - during heat transfer the thermal energy flow always involves the
polymer matrix. The irregular crystalline structure generated by the presence of
higher amounts of xGnP does not affect the electrical conductivity as long as a
continuous path is formed by the nanoplatelets, but is not beneficial for achieving
high thermal conductivity.

The possibility of a percolation threshold for thermal conductivity was
intriguing, since previous studies on polymer/CNT composites indicate that
thermal conductivity does not show a clear percolation.“ '1

As described in the introduction to this chapter, the heat transfer in
polymer/CNT composites is governed by the relatively small contrast between
the thermal conductivity of the matrix and that of the nanotubes. Numerical
simulations4 indicated that if the contrast between thermal conductivity of the
matrix and thermal conductivity of the CNT increases by a factor of 1000, the

heat flux at contact increases by two orders of magnitude. Even sharper relative

147

increases were predicted for larger thermal conductivity ratios, and strong
percolation threshold in conductive transport was predicted. Our results indicate
that the either the contrast between the thermal conductivity of exfoliated
graphite nanoplatelets and the thermal conductivity of the PLA is higher than the
expected 10“, or the interfacial thermal resistances plays a more important role
as the xGnP loadings increase. The above-mentioned simulation results indicate
that in order to achieve a percolation threshold for thermal conductivity, the
difference between the thermal conductivity of the polymer and that of the
nanoplatelets should be of the order of 107. Such a high contrast is not very
probable since the thermal conductivity of the PLA matrix is about 0.20 W/mK,
and the conductivity of xGnP cannot be exceeding the thermal conductivity of
individual graphene sheets.1 We believe that the interfacial thermal resistance
between graphite nanoplatelets is lower than the polymer-xGnP interfacial heat
resistance, and the more nanoplatelets in contact with each other, the higher the
conductivity of the composite. Larger improvement in thermal conductivity was
not achieved because as the amount of xGnP was large enough to allow
continuous contact between nanoplatelets, the crystalline structure of the
polymer matrix interfered with the ability of the platelets to form a conductive
path. When xGnP-1 was dispersed in immiscible PLA/PS blends, the phase
separation between PLA and PS directed the xGnP towards forming a

continuous path (Figure 5.5).

148

 

Figure 5.5 Optical micrographs of a) PLA L21OS containing 10 % xGnP-1 and
b) composite sample containing 70 % PLA L21OS, 20 % PS, and
10 % xGnP-1

The thermal conductivity of PLAIPSIxGnP-1 systems showed the potential of
the polymer blend to lead to better thermal transfer. As shown in Figure 5.6, the
thermal conductivity of PLA is approximately 0.20 W/mK, PS has a lower
conductivity of 0.13 W/mK (due to its amorphous nature), while a mixture of 90 %
wt PLA and 10 % wt PS has intermediate thermal conductivity of 0.17 W/mK.
Addition of xGnP-1 to the PLA/PS blends helps improving the thermal
conductivity. For example (Figure 2.6.10), PLA containing 10 % xGnP-1 is less
conductive than a composite containing 80 % PLA,.10 % PS, and 10 % xGnP-1.

The same trend can be seen for composites containing 20 % xGnP-1.

149

 

 

 

 

 

100/0/0 0/100/0 90/10/0 90/0/10 80/10/10 60/30/10 80/0/20 70/10/20 70/0/30

 

 

composite sample

 

Figure 5.6 Thermal conductivity of PLAIPSIxGnP-1 composites

The results obtained for the thermal conductivity of PLAIPSIxGnP-1
composites show that taking advantage of preferential localization of xGnP in a
blend of immiscible polymers can lead to improvement in thermal conductivity.
However, for PLA/PS systems analyzed, this approach is in the detriment of the
mechanical properties. Figure 5.7 shows that the modulus of the polylactides
used in these experiments is 3.19 GPa, and the modulus of PS is 2.69 GPa. The
modulus of PLA samples containing 1 % wt xGnP-1 is 30 % higher than that of
pure PLA. PLA containing 1 % wt xGnP-1 and increasing amounts of PS show a
decrease in modulus for 10 %, 20 %, and 30 % wt PS. The modulus of PLA/PS
blends containing 1 % xGnP-1 increases again after phase inversion, when the

system starts to be dominated by the PS.

150

 

 

 

  

 

 

 

 

 

5.00 7 K \
reference PLA/PS containing 1 % xGnP-1

EL“? 4 00 _ polymers 3.47
g ' 3.19 3'28
in 2.69
.3. 3.00 7
3
'o
o
E 2.00
to
on
E
9 1.00 7
a)

0.00 *

PLA PS PLA 10%PS 20%PS 30%PS 40%PS 50%PS
Composite sample

 

 

Figure 5.7 Storage moduli of PLA, PS, and PLAIPSIxGnP-1 composites
containing 1 % xGnP-1

Conclusions

Exfoliated graphite nanoplatelets can impart thermal conductivity to PLA. The
improvement was as high as 300 % for 30 % xGnP-1 in PLA L21OS, and over
500 % for 15 % xGnP-15 in the same matrix. The results indicate that the
improvement in thermal conductivity is limited by the interfacial heat transfer
between the polymer and the graphite nanoplatelets. Larger graphite
nanoplatelets are more efficient in improving thermal conductivity, as the
contribution of the polymer/graphite interfacial layer to thermal conductivity is
inversely proportional to the size of the platelets.

As the amount of xGnP in the PLA matrix increases, regardless of platelets

size, the nucleating effect of xGnP on the PLA determines the rapid formation of

151

crystalline structures that are more numerous and more irregular, increasing the
overall disorder in the polymer matrix, which limits the improvement in thermal
conductivity.

Better improvement in thermal conductivity could be achieved by using
biphasic polymer blends, by directing the graphite nanoplatelets either in one of
the polymers, or at the interface between the immiscible polymers. In the
PLA/PS system studied, the improvement in thermal conductivity was in the
detriment of mechanical properties, which showed a typical phase inversion
behavior, as the system was dominated either by the semicrystalline PLA, or by

the amorphous PS.

152

References

10.

11.

Balandin, A. A.; Ghosh, S.; Bao, W.; Calizo, l.; Teweldebrhan, D.; Miao,
F.; Lau, C. N., Superior Thermal Conductivity of Single-Layer Graphene
Nano Letters 2008, ASAP.

Winey, K. l.; Kashiwagi, T.; Mu, M. F., Improving electrical conductivity
and thermal properties of polymers by the addition of carbon nanotubes as
fillers. Mrs Bulletin 2007, 32, (4), 348.

Winey, K. l.; Vaia, R. A., Polymer nanocomposites. MRS Bulletin 2007,
32, (4), 314.

Shenogina, N.; Shenogin, S.; Xue, L.; Keblinski, P., On the lack of thermal
percolation in carbon nanotube composites. Applied Physics Letters 2005,
87, (13), 133106.

Huxtable, S. T.; Cahill, D. G.; Shenogin, S.; Xue, L. P.; Ozisik, R.; Barone,
P.; Usrey, M.; Strano, M. S.; Siddons, G.; Shim, M.; Keblinski, P.,
lnterfacial heat flow in carbon nanotube suspensions. Nature Materials
2003, 2, (11), 731.

Hull, 0.; Clyne, T. W., An Introduction to Composite Materials. Second
Edition, Cambridge University Press, 1996.

Gubbels, F.; Blacher, S.; Vanlathem, E.; Jerome, R.; Deltour, R.; Brouers,
F .; Teyssie, P., Design of Electrical Conductive Composites - Key Role of
the Morphology on the Electrical-Properties of Carbon-Black Filled
Polymer Blends. Macromolecules 1995, 28, (5), 1559.

Gubbels, F.; Jerome, R.; Teyssie, P.; Vanlathem, E.; Deltour, R.;
Calderone, A.; Parente, V.; Bredas, J. L., Selective Localization of
Carbon-Black in lmmiscible Polymer Blends - a Useful Tool to Design
Electrical Conductive Composites. Macromolecules 1994, 27, (7), 1972.

Wan, Y.; Fang, Y.; Hu, Z. L.; Wu, 0., Electrically conductive poly(DL-
lactide)/chitosan/polypyrrole complexes. Macromolecular Rapid
Communications 2006, 27, (12), 948.

Miloaga, D. G.; Hosein, H. A. A.; Misra, M.; Drzal, L. T., Crystallization of
poly(3-hydroxybutyrate) by exfoliated graphite nanoplatelets. Journal of
Applied Polymer Science 2007, 106, (4), 2548.

Biercuk, M. J.; Llaguno, M. C.; Radosavljevic, M.; Hyun, J. K.; Johnson, A.

T.; Fischer, J. E., Carbon nanotube composites for thermal management.
Applied Physics Letters 2002, 80, (15), 2767.

153

CHAPTER 6

SCANNING PROBE THERMAL ANALYSIS OF POLYLACTIC ACIDI
EXFOLIATED GRAPHITE NANOPLATELET NANOCOMPOSITES

Scanning probe thermal analysis, also known as nanoscale thermal analysis
(nano-TA), is introduced here as an important nanoscale characterization tool for
PNC. The technique was used in this research for the analysis of the nanoscale
structure of PLA/xGnP nanocomposites. The first part of this chapter presents
the investigation of calibration standards to be used for the analysis of polymer
nanocomposites in which the polymer matrix melts under 200 °C, and
experiments conducted to determine the effect of heating rates on the measured
melting temperature of a variety of polymers. The second part, which constitutes
work published in 20081 in Journal of Biobased Materials and Bioenergy,
presents the use of nano-TA together with conventional therrnomechanical
analyses to offer insight on the effect of xGnP on the crystallization behavior of
PLA, and to explore the relationship between the polymers molecular weight and
the dispersion of xGnP in the matrix. Specifically, the technique was used to
monitor cold crystallization phenomena, and to assess the dispersion of

nano-size fillers in the polymer matrix.

154

Calibration standards and testing conditions for the study of PHB/xGnP

and PLA/xGnP nanocomposites

Introduction

Nano-TA is a local thermal analysis technique utilizing a novel micromachined
scanning probe microscopy tip with an integrated heater to monitor thermal
behavior of materials at nanoscale dimensions. The measurements consist in
positioning the tip at a fixed location, increasing the tip temperature and monitor
the deflection, to detect transitions such as melting temperature (Tm) and glass
transition temperature (T9).

The spatial resolution for imaging with this system is reported to be one
hundred times better than its predecessor, micro-TA (pTA), primarily due to
advances in probe technology. The new probes have increased tip height and
sharpness, enhanced cantilever geometry for laser alignment and relatively low
spring constants, which are important features for imaging soft materials such as
polymers in the contact mode without causing significant damage to the surface
of the specimen.

For polymers and PNC, while DSC gives quantitative information regarding
the dependence of the degree of crystallinity on cooling rates, nano-TA offers the
unique advantages of in-situ observation of the polymers morphology under the
same conditions, providing valuable information for practical applications.

A special nano-TA probe containing an embedded miniature heater is used in

lieu of the conventional AFM probe (see Figure 6.1 a).

155

 

20 um

 

 

 

 

Figure 6.1 Scanning Electron Micrographs of a first generation nano-TA probe
showing a) entire cantilever; b) the 1pm tip) and c) a temperature
map of the probe (Images courtesy of Anasys Instruments, CA)

Figure 6.2 shows representative AFM images of a PLA test area before and
after nano-TA. These images attest to the high quality of surface

characterization obtainable with the nano-TA probe.

 

 

 

 

 

Figure 6.2 AFM images of a PLA surface before nano-TA a) height image, and
b) deflection image and C) after nano-TA, showing the indentation
formed when the sample melts (Scans are all 3 pm2 and collected
at a scan rate of 2 Hz.)

156

Figure 6.2 shows details of the morphology of the PLA sample. Although it
appears that an area of approximately 2 pm2 was affected by the nano-TA
measurement, the actual size of the indent is on the order of a couple hundred
nanometers. The area shown in Figure 6.2b is larger due to polymer that
recrystallized after the tip was withdrawn, and the region was brought back to
room temperature.

This type of probe stems from the work of Dlnwiddie et al.,2 who combined
thermal conductivity measurements with AFM mapping topography. Others have
since contributed to this technology.” Together with specially designed
hardware and software, it is possible to obtain nanoscale resolution of a
material’s surface using either contact or tapping mode AFM, while
simultaneously applying heat locally through the probe tip for thermo-mechanical
investigations. The nano-TA measurement is analogous to TMA, which records
changes in the physical dimensions of a sample as a function of temperature
and/or time.

Nano-TA data of cantilever deflection as a function of probe tip temperature
for PHB is shown in Figure 6.3. The initial upward deflection of the probe
cantilever is the results of the thermal expansion of the sample and the sudden

sharp downward deflection is indicative of the melting transition.

157

 

Deflection (V)
N (a)

 

 

0 I I I I I IJ‘F I I
-120 2/60 80 100 120 140 160 180 200

 

Temperatue (°C)

 

 

 

Figure 6.3 Example of nano-TA thermal profile for PHB

Unlike pure elements, which have a well defined melting temperature,
polymers melt over a broad temperature range. Polymer melt profiles are
routinely measured using DSC and the melt temperature is typically assigned as
the peak temperature of the endothermic melting event. Due to molecular weight
distribution considerations, crystalline structure and size effects, the onset of
melting generally occurs well before the peak endothermic temperature.
Consequently, it is often useful to define the extrapolated onset melting
temperature as the point where the baseline intersects the tangent through the
linear section of the first order endothermic event before the peak temperature is
attained. The difference between the temperature corresponding to the
extrapolated onset of the melting transition and the temperature of peak recorded

for the melting endotherm can be substantial.

158

 

As is the case with differential scanning calorimeters, the nano-TA unit
requires detailed temperature calibration procedures using known standards, as
well as an understanding of the precision of the temperature measurement.
Since the technique is new, it was imperative to evaluate calibration standards
that would allow accurate measurements within the desired temperature range.
The focus of this work being nanocomposites from PHB and PLA, the desired
temperature range was room temperature to 200 °C (~ 20 °C above the melting
temperature of each of the polymer matrices).

A good calibration standard should have a number of desirable properties,
such as a well-documented melting temperature, availability in a suitable form for
analysis, homogeneity, stability, low toxicity, and traceability to a national
reference laboratory.7 The calibration standards used for micro-TA are polymeric
materials since these do not interact with the platinum sensor of the micro-TA
probe. It is also essential to use pure, crystalline polymers with a single
polymorphic form.

PCL, PE, and PET possess the requisite properties to serve as standards for
calibrating the temperature response of the nano-TA probe. Pro-fax 6301, a PP
homopolymer, was also investigated for the use as calibration standard.

The effect of heating rates on polymeric thermal melt behavior was assessed

using the nano-TA instrument at ramp rates from 60 to 600 °C per minute.

159

Experimental details

Materials

Poly(§-caprolactone), polyethylene, and poly(ethylene terephthalate
calibration standards were provided by Anasys Instruments, CA and used as
received. Pro-fax 6301 was obtained from Basell Polyolefins and was used

without further purification.

Sample preparation

PCL, PE, and PET calibration standards were used as received. Thin films of
Pro-fax 6301 were obtained in a Mettler Toledo FP82 hot stage by heating at
20 °Clmin a small amount of polymer sandwiched between a glass slide and a
cover slip. These samples were obtained after nonisothermal crystallization
upon cooling ~20 °Clmin to room temperature from 190 °C. To ensure complete

melting, the samples were held isotherrnally above 190°C for five minutes.

Methods

8) Differential scanning calorimetry

DSC measurements were made using a 2920 DSC (TA Instruments, New
Castle, DE) operated in the standard mode with a nitrogen gas purge. The unit
was calibrated using an Indium standard. For each run, a fresh sample (5-10
mg) of polymer was encapsulated in crimped aluminum pans. Three different

samples were tested for each polymer. Samples were tested at a heating rate of

160

10 °Clmin from room temperature to 100 °C, 150 °C, 280 °C, and 250 °C
respectively.

The extrapolated onset temperature, Tmo, was assigned by the intersection of
a tangent constructed to the baseline with a tangent through the linear portion of
the first order melt transition. The manufacturer's software was employed to
construct the tangents for determination of the extrapolated onset melting point,
Tmp.. The peak melt temperature was determined by using the vendor’s software
tools to locate the temperature where the material was at its peak endothermic

response.

b) Scanning probe thermal analysis

Tests were performed using a nano-TA instrument (Anasys Instruments, CA)
in conjunction with a Nanoscope IV (Veeco Instruments, CA) scanning probe
microscope. A silicon micromachined thermal probe, with a cantilever thickness
of approximately 2 pm and having a resistivity of about 1 k0 was used for all the
tests. Heating was made possible through a high resistance region within the
probe, and controlled by varying the voltage applied through the nano-TA
electronics. The maximum temperature attainable for this particular probe was
approximately 250 °C. The probe temperature was initially calibrated with PCL
(60 °C), PE (130 °C), and PET (238 °C). AFM imaging was done in the contact
mode at a scan rate of 2 Hz for scan sizes of 3 pm2 to locate and select an

appropriate area of the surface for nano-TA testing.

161

Results and discussion

PCL PE and PET as calibraflon standards

 

 

The DSC results for PCL, PE, and PET are presented in Table 6.1 and
Figure 6.4. For PCL and PE, the Tmp values determined by DSC were in close
agreement with the literature melting temperatures, T,,,(lit). However, for PET,
the Tmo was in better agreement with the Tm(lit) value. The Tmp of PET from DSC
was measured to be ~17 °C higher than the Tm (lit) although the cause for this
discrepancy is not known.

The nano-TA melt temperatures for PCL and PET (Table 6.1) closely
correspond to the Tm values determined by the DSC. This was not the case for
PE, which showed a large variability in the nano-TA results compared to the

other polymers used as calibration standards.

162

Table 6.1

Melting temperatures of PCL, PE, PET and Pro-fax 6301 obtained
from DSC and nano-TA experiments

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample DSC Nano-TA Literature
value
Tmo. [°C] Tmp [°C] Tm [°C] Tm ("0 [°C]
PCL
PCL1 54.27 62.01 55.26
PCL2 54.69 61.53 56.41
PCL3 54.94 61 .73 55.84 60
PCL Average 54.63 61 .76 55.82
Stdev. 0.34 0.24 0.60 -
PE
PE1 117.99 130.04 126.95
PE2 115.53 129.40 138.04
PE3 118.33 130.26 140.41 130
PE Average 1 17.28 129.50 135.14
Stdev. 1.53 0.45 7.19 -
PET
PET1 238.03 256.91 235.56
PET2 236.01 253.84 230.42 238
PET3 232.37 253.94 235.83
PET Average 235.47 254.90 233.94
Stdev. 2.87 1 .74 3.05 -
Pro-fax
Pro-fax1 147.58 162.61 142.21
Pro-fax2 147.84 163.23 139.54 160
Pro-fax3 147.92 162.81 142.54
Pro-fax Average 147.78 . 162.88 141 .43
Stdev. 0.18 0.32 1.65 -

 

163

 

 

 

 

Heat flow (ng)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-15 - —PCL3
3)
-20 J 0
Temperature( C)
10 7

123
E

3

O
c
‘0'

0
I

b —PE3
-30 7 )
Temperature (°C)
2 1

2'7

E

3

E:

*6

0

I

 

 

 

 

 

 

Terrperature (°C)

 

 

Figure 6.4 DSC curves showing the melting behavior of a) PCL; b) PE and
C)PET

164

Figure 6.5 shows a summary of the Tmo, Tmp, Tm(lit), and Tm for the three
calibration standards. It should be noted that the standard deviations of the Tmo
and Tmp values obtained by DSC were smaller for PCL, followed by PE, then
PET. Except for PE, the standard deviations calculated from the nano-TA results
were comparable with those for DSC data collected on the same material. This
indicates that nano-TA has the potential of being used as an effective thermal
Characterization tool, leading to reproducible results comparable to a widely-used
and accepted calorimetlic technique. The exact cause of the variability of the
nano-TA results on PE was not determined, but generally the inherent
characteristics of polymer samples (e.g. degree of crystallinity, molecular weight
distribution), or even the variability in the response of the nano-TA probe as a
function of temperature, should be considered as contributing factors.
Additionally, it must be considered that properties of surfaces may be dissimilar
to the bulk properties of a material. The ability to measure melt properties with
nano-TA may not reflect the bulk material properties. Based on these results,
Pro-fax 6301 was investigated as an alternative calibration standard to PE. Both
the DSC and the nano-TA results for this material were very reproducible,
showing small standard deviations (Table 6.1). Hence, Pro-fax 6301 has

suitable thermal behavior to be a viable standard for nano-TA calibration.

165

 

    
     
   

300

 

 

I Tmo

a a Tm(na no-TA)
“g; 200 Tmllit)

.5

g 150 nu... :::::t
o. IIIIIIIIIIIII
§ 100

50
0 am:

PCL PE PET Pro-fax
Calibration standard

 

 

 

Figure 6.5 Summary of DSC results showing Tm, Tmp, Tm(lit), and Tm(nTA)
values for PCL, PE, PET and Pro-fax

Nano-TA heating rate effects on polymer melt behavior

The effect of nano-TA heating rate on the melt behavior of Profax 6301, PE,
PHB, and PET was evaluated at temperature ramp rates of 60 to 600 °Clmin.
The tip was placed on a random location on the surface and three tests were
performed for each heating rate. As soon as the melting transition was
observed, the tip was quickly withdrawn from the surface.

The nano-TA heating rate data are presented in Figure 6.6 and Table 6.2.
The Tm values show good reproducibility, and the melting transitions are obvious.
In general, the Tm values were not significantly affected by the heating rate,
allowing for reproducible results even for scan rates as high as 600 °Clmin. It

was observed that as heating rates increased, the sharpness of the melting

166

transitions decreased slightly The rates of thermal expansion, as reflected in the
slope of cantilever deflection versus temperature curve, were also observed to be
slightly higher for slow heating rates (10 and 20 °Clmin) compared to faster rates
(> 30 °Clmin). At slower ramp rates, the probe is in contact with the sample for
longer time periods, giving more time for the sample to undergo thermal

expansion compared to faster heating rates.

 

 

 

 

 

 

 

 

 

 

 

 

Table 6.2 Melting temperatures recorded by nano-TA using scan rates from
60 to 600 °Clmin
Heating Rate Tm (Nano-TA) (°C)
(°Clmin) PE PET Pro-fax
60 129.6 226.3 140.8
120 130.0 228.4 140.2
180 135.1 231.7 140.9
240 128.2 230.1 141.8
300 133.6 229.0 143.6
360 129.4 228.7 140.2
420 130.5 226.8 139.6
480 130.7 227.6 141.6
540 132.1 228.1 144.4
600 126.4 226.2 140.2

 

 

 

 

 

167

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

7 —10 9
g —8 —7

,0 --6 5

>

7; 4 ~—4 —3

.g 3 ——2 —1

8 2

‘5 1

D
0 I I I
71%) 110 140 170 200 230 260
-227'L)

8)
Temperature (°C)
8-
—10 9
6- —8 —7

E 72 .2.

5:3 4' —-—2 —1

8

1.: 2‘

o

o
o- | I I I
220 110 140 170 200 230 260

b)
Temperature (°C)
10“—10 9
87—8

,0 —6

a 6‘ ——4

.8 4~--2

0)

D 0 / I j
-220 50 80 110 140 170 200 23 260
-4— C)

Temperature (°C)

 

Figure 6.6 Nano-TA results showing the effect of heating rate on the melt
temperature for a) Pro-fax 6301, b) PE and c) PET

168

Conclusions

PCL, PE, and PET were proved to be efficient calibration standards for nano-
TA experiments requiring heating up to 200 °C. The results of this study are
important since nano-TA is not an established method for investigation of PNC
and a good understanding of the methods advantages and limitations is requires
for before subsequent use in nanoscale characterizations.

When used in conjunction with AFM, nano-TA can provide thermal data
related to localized areas of a sample. In this respect, nano-TA provides
information specific to the surfaces of materials on the nanoscale which are not
obtainable by bulk investigation methods such as DSC. Nano-TA can
measure melt temperatures (T,n) of polymers at very fast heating rates. In this
study, varying the heating rate from 60 to 600 °Clmin gave consistent and
reproducible identification of the melting points, qualifying nano-TA as a viable
method to measure the melt temperature of many polymeric materials. However,
it should be noted that heat loss still occurs both within the probe and to the
atmosphere. It is difficult to ascertain the extent of this loss, and also difficult to
determine how this affects the Tm values measured. The standard deviations of
Tm values obtained via nano-TA were always larger compared to the
corresponding values from DSC measurements, but this variability may be due in
part to local surface properties of the materials such as roughness, as well as
atmospheric conditions like humidity.

Finally, a note of caution was raised regarding potential artifacts resulting

from sample preparation. Since the nano-TA interacts with only the first few

169

nanometers of the sample surface, consistent methods of sample preparation
must be applied in order to insure that the results are representative of the

sample and not a local artifact resulting from surface preparation.

Experimental investigation of PLA/xGnP

Introduction

Scanning probe thermal analysis was further employed for the investigation of
PLAIxGnP-1 nanocomposites prepared by solution mixing and compression
molding. This preparation technique was shown (Chapter 4) to lead to the best
dispersion of xGnP-1 in the polylactides matrices. The purposes of the
nanoscale investigations presented here were to relate modifications in polymers’
morphology induced by the xGnP to changes in the mechanical properties of the
composites, and to establish if the high nucleating effect of xGnP is a limiting

factor in improving the properties of resultant nanocomposites.

Experimental details

Materials
Poly(L-Iactic acid) was purchased from Boehinger-lngleheim Chemicals

(Petersburg, VA). Two different molecular weight PLA samples were used in this

170

study: RESOMER® LZOQS (intrinsic viscosity 3.0 dl/g for 0.1 % solution in
chloroform at 25 °C), and RESOMER® L2108 (intrinsic viscosity 3.8 dl/g for 0.1
% solution in chloroform at 25 °C). Methylene chloride and 2—propanol were
purchased from Sigma Aldrich. Exfoliated graphite nanoplatelets having an
average size of 1 um (xGnP-1) was produced in-house from GrathuardTM 160-
50A acid intercalated graphite produced by Graftech (Cleveland, OH), as

described in Chapter 2.

Mtion of PLA/xGnP-1 composites

PLA composites with xGnP-1 were prepared by solution mixing and
compression molding. The xGnP-1 loading varied from 0 wt % to 9 wt % for both
PLA samples (L2098 and 2103). For each xGnP-1 loading, 1 g of PLA/xGnP-1
mixture was prepared as follows: appropriate amounts of xGnP-1 were dispersed
in 20 mL methylene chloride by sonication (5 min, 10 W) using a VirSonic 100
Cell Disrupter from Virtis (Gardiner, NY). Corresponding amounts of PLA were
further dissolved in the xGnP-1/ methylene chloride dispersions, and the
obtained PLA/xGnP-1 mixtures were precipitated using 2-propanol. The resulted
PLA/xGnP-1 materials were kept at room temperature overnight, and then dried
in a vacuum oven at 80 °C for 20 h to ensure the complete evaporation of the
organic solvents. The dried PLA/xGnP-1 mixtures were compression molded at
200 °C for 2 min, under an applied pressure of 18,000 psi, using a Carver
laboratory press (Carver Inc., Wabash, IN). The press” platens were pre-heated

to 200 °C prior to compression molding.

171

Methods

a) Dynamic Mechanical Analysis

The storage moduli of PLA, and PLA/xGnP-1 composites were evaluated
using a 2980 Dynamic Mechanical Analyzer (TA Instruments, New Castle, DE).
The samples (approximately 5.50 mm x 16 mm x 0.10 mm) were cut from the
compression molded films, and tested in tension during heating from room
temperature to 120 °C. The reported values are for storage modulus at room
temperature, and represent the average of three measurements per xGnP-1

concentration in the PLA.

b) Differential Scanning Calorimetry

The melting and crystallization behavior of PLA and PLA/xGnP-1 composites
were investigated using a using a 2920 Modulated DSC (TA instruments, New
Castle, DE). The runs were performed under nitrogen flow, after calibrating the
DSC instrument using an Indium standard. For each run, a fresh sample (6-10
mg) of polymer/ composite was placed in a seal aluminum pan and subjected to
two heatingl cooling cycles. During the first cycle, incorporated in the DSC
method to ensure the erasure of samples’ thermal history, the samples were
heated (10 °Clmin) from room temperature to 200 °C, kept isothermal at 200 °C
for 10 min, cooled (1O °Clmin) to 0 °C, and equilibrated at 0 °C. The second
heating/cooling cycle immediately followed and involved heating (3 °Clmin) to
200 °C, keeping the samples isothermal at 200 °C, then cooling (3 °C) to room

temperature. Except for the data in Figure 5, where the first heating cycle was

172

 

presented, all data reported and analyzed refer to the second heatingl cooling

cycle.

0) Optical microscopy

The samples for optical microscopy were prepared as thin polymer films by
heating very thin sections of PLA/xGnP-1 compression molded composites in a
Mettler Toledo FP82 hot stage. The polymers were placed on glass slides,
covered with glass cover slips, and heated at a rate of 10 °Clminute from room
temperature to 200 °C, kept isothermal for ten minutes, then allowed to crystallize
upon cooling to room temperature. The morphology of PLA and the size of the
spherulites were observed using an Olympus BH2 optical microscope equipped

with a SpotTM camera.

d) Scanning probe thermal analysis

The tests were performed using a nano-TA instrument (Anasys Instruments,
CA) in conjunction with a Nanoscope” IV instrument (Veeco, Santa Barbara,
CA) equipped with a J scanner, as described in the first part of this chapter. The
maximum temperature attainable for this particular probe was approximately
250 °C. The probe temperature was initially calibrated using polymer standards:
poly(s-caprolactone) (mp. 60 °C), polyethylene (mp. 130 °C), and polyethylene
terephthalate (m.p.238 °C), as described in the previous section.

The tested surfaces were sections of the compression molded composite

films prepared as described above. In order to select areas on the surfaces that

173

were appropriate for the nano-TA testing, initial 3 pm2 scans were collected in
contact mode without heating the probe, using a 3 Hz scan rate. The scan size
was reduced to 1 nmz, and the probe was heated at a rate of 10 °C/sec from 25
°C to an end temperature of 180 °C. For all the tests, the delay was 0.1 sec, the
hold time was 0.1 sec, and the probe cooling rate was 20 °Clmin. Data were
collected at a rate of five readings per second, and the reported results represent

the average of five measurements per composite sample.

Results and discussion

Mechanical properties of P4LA/x9nP composites

The storage moduli of PLA L21OS/xGnP-1 composites, as determined by
DMA, increased with increasing amounts of xGnP-1 in the polymer matrix. The
improvement was consistently greater than 50-60 % for 3-9 % wt xGnP-1 in
PLA L21OS. The same amounts of xGnP-1 did not lead to similar improvement
of the storage modulus when incorporated in the lower molecular weight
polylactic acid, PLA LZOQS (Figure 6.7). In the PLA L209$ composites, the
storage modulus improvement was greater than 50 % only for composites with

6-7 wt % xGnP.

174

 

I PLA L2098/ xGnP-1
7 - E PLA L210/xGnP-1

Storage Modulus (GPa)

 

 

0 1 3 5 7 9

 

 

xGnP-1 (% wt)

 

Figure 6.7 Storage moduli of PLA/ xGnP-1 composites

Cgstailization of PLA/xGnP composites

Polylactides are semicrystalline polymers, and their mechanical properties, as
well as the mechanical properties of the composites for which they serve as
matrices, are expected to depend on crystallinity and morphology. Since the
composite samples were prepared using compression molding, a dynamic
processing method, the nonisothermal crystallization from the melt was analyzed
in order to elucidate the differences observed in storage moduli. The DSC
analysis of PLA L2098, PLA L21OS, and their composites with xGnP-1 revealed
that the crystallization behavior of both PLA samples was significantly affected by
xGnP-1. Low amounts of xGnP-1 (up to 1 % wt) lead to increased melt
crystallization temperatures (Figure 6.8, and Table 1), which indicated a
nucleating effect of the nanoplatelets, similar to the nucleating effect observed

recently in PHB/xGnP-1, and PP/xGnP-1 compositesg‘m. The nonisotherrnai

175

crystallization peaks for PLA L2098 and PLA L21OS containing 0.01 % wt xGnP-
1 were narrower than the peaks recorded for the neat polymer samples, which

indicated that the crystallization process was faster in the presence of xGnP-1.

 

 

 

 

 

 

0.6 % wt xGnP-1: a)
'27
E
X
3
c
76
o
:I:
90 100 110 120 130 140 150
Temperature (°C)
0-6 ‘ % wt xGnP-1: b)

E

E

X

3

G:

‘6

o

I

90 100 110 120 130 140 150
Terrperature (°C)

 

 

 

Figure 6.8 DSC analysis of the nonisothermal crystallization from the melt of
(a) PLA L2098 and (b) PLA 2103 and their composites with
xGnP-1

176

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

177

 

 

 

 

 

 

 

 

 

 

8.8 8.8 888 88 FF 8.8 88. 8F ”8.8: 88F 8.8 8
8.8 8.8 888 8.8 8F.8F ”FFdFF 8.8FF 8 F8 8
8.8 8.8 N F. FNF 8.8 8.8F u8.FFF 8.8 F 8.8 F
8.8 8.8 888 N F. F FF 8.8 Fm. 8F 8.8FF 8.8FF 8.8 8
8 F. 8 88 88F 8.83 8.8 8. 8F 8. 8F 8.8FF 8 F8 8
8.8 8.8 8.8 F 8.8 Fm. 8F 8.8: 88F 8.8 v
8.8 8.8 88F F 8 F8 88. 3: ”NSF 88F 8.8 8
8.8 8.8 FF.oF F ”8. FNF 8.8 8. 8F 8.8: 8.8FF 8.8 N
8.8 8.8 8.8 8.8 8.8FF Fo.8FF 8.8 F
8.8 8.8 888 8.8 8F8FF 882 8.8 8.8
8. FF 8.8 8.2: 8. F8 8.8FF 8.8 F 8.8 8
780588.. <._n_
8.8 8.8 Fm. F FF 8.8 88. 8F 8: 8.8 8
8.8 8.8 FvFNF 8 F8 FF N F. FF. 8. 8F ” 88 F 8.8 F 8.8 m
8.8 8.8 8.8 ”88.: F8 88. 8F ”FYFFF «EFF 8.8 F
8.8 8.8 8.8 FF F F. FF. 8.8 F ”FNFF F 8.8 F 8.8 8
8.8 8.8 8.8F ”8.~FF 8.8 88. 8F 8.~FF 8.8 m
8.8 8.8 8.8F 8. F FF 8.8 8. 8F 8.8 F 8.8 v
8.8 8.8 8.88 8. F FF 8.8 88. 8F 88: F~8FF F F8 8
8.8 88. FF. 8.F~F 88F F 8.8 8.8FF FNVFF 8.8 N
8.8 8.8 8.8 F 882 8.8 8.8 F 88F F 8. 8 F
8.8 8 F8 8.88 8.8 8.8F 8.8 F 88F F 8.8 8.8
8.8 8.8 FF.8FF 882 8.8 8.2: 8F8FF 8.8 8
780888 Sn.
3°C 60V . 8d. 4008 8% c; as
.x late an era SF SF uF 88
8:88:80

F-n_c0x\<._n_ F0 :88 9: E9: co_FmN___mFmbo 8.505028: ncm 958E 8 £8.88 own .6 meEzw «.8 Sam...

In nanocomposites containing xGnPa' 9' narrower crystallization peaks were
also correlated with smaller crystallites formed by the polymers in the presence
of xGnP-1, as shown also in Chapter 2 of this dissertation.

Figure 6.9 shows an example of the effect of xGnP-1 on the size of the
crystalline structures formed by PLA L21OS with nonisothermal cooling from the
melt. The neat PLA forms spherulitic structures 20 — 100 pm in diameter (Figure
6.9,a). The addition of 0.01 % wt xGnP-1 lead to formation of much smaller
crystalline structures (Figure 6.9,b), while crystalline structures were not
detectable using optical microscopy when the amounts of xGnP-1 exceeded 1 wt
%. Similar nucleating effects of nanosize reinforcements, leading to higher
temperatures of crystallization from the melt, and reduction in spherulitic size,

10, 11’ and

were reported for polylactic nanocomposites with carbon nanotubes
layered silicates.10 Similar crystallization behavior was reported for
nanocomposites obtained from other nanocomposites of semicrystalline
polymers, such as nylon-6/clay‘2, polypropylene/clay“, and
polypropylene/carbon nanotubes. 14' 15

For 1 % wt xGnP-1, a clear difference was observed between PLA LZOQS and
PLA L21OS: for the low molecular weight PLA sample, a bimodal nonisothermal

crystallization curve was observed, which could be related to dispersion issues

leading to hindrance of crystal growth.16'17

178

 

Figure 6.9 Optical micrographs showing spherulites of (a) pure PLA L21OS; (b)
PLA L21OS/ 0.01 % % wt xGnP-1 and (c) PLA L21OS/ 5 % wt
xGnP-1 formed during nonisothermal crystallization from the melt.
The scale bar is 20 pm.

No significant differences in the melt crystallization behavior were observed
between the composites of the two different molecular weight polylactides
containing amounts of xGnP—1 higher than 1 % wt. The melt crystallization peaks
recorded for composites containing more than 1 % wt xGnP-1 were wider and

occurred at lower temperatures than the peaks recorded for low amounts of

179

xGnP-1, regardless of polymers molecular weight. Bimodal crystallization
curves were further observed for increased amounts of xGnP-1 in both matrices,
indicating possible agglomeration of the platelets and creation of polymer-rich
regions. The polylactides crystallized heterogeneously, at higher rates, in the
vicinity of xGnP-1, while in the polymer-reach regions, the crystallization was
homogeneous and similar to the neat polymers’.

The DSC analysis of neat PLA L209$ and PLA 2108 samples also revealed
that the neat polymers undergo cold crystallization, upon heating from room
temperature. The cold crystallization peak temperatures were recorded between
93 — 106 °C (Table 6.4), depending on the heating rate, as will be discussed
further in this paper. The DSC analysis did not detect oold crystallization after

addition of xGnP-1 to either one of the polymer matrices.

180

Table 6.4

Summary of the DSC analysis of cold crystallization of PLA L209$
and PLA L21OS for different heating rates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ticc (°C) ch (°C) AHcc (°C)
Heating
rate 3 5 10 3 5 10 3 5 10
(°Clmin)
PLA L209s
Test1 87.92 89.61 95.65 94.27 98.61 105.03 22.46 27.15 28.88
Test2 86.04 91.19 93.36 94.27 99.07 105.16 23.47 27.05 27.10
Test3 88.12 89.17 93.60 94.52 98.58 104.03 24.87 22.78 27.46
Average 87.36 89.99 94.20 94.35 98.75 104.74 23.60 25.66 27.81
332$)? 0.94 0.87 1.03 0.12 0.22 0.50 0.99 2.04 0.77
PLA L21os
Test1 80.62 82.62 88.42 93.43 95.26 102.94 7.231 11.31 13.75
Test2 81.89 86.01 89.98 94.10 97.17 104.24 14.31 16.61 17.30
Test3 85.70 83.93 92.24 94.82 96.60 103.06 12.03 13.00 9.836
Average 82.74 84.19 90.21 94.12 96.34 103.41 11.19 13.64 13.63
3:11:30? 2.16 1.40 1.57 0.57 0.80 0.59 2.95 2.21 3.05
Melmf PLA/xGnP composites

 

 

The DSC analysis indicated that neat PLA L209$ and PLA L21OS melt at

179-180 °C (Table 6.3). For both polymer composites with xGnP-1, bimodal

melting peaks were recorded. Additional peaks (shoulders) were observed at

181-182 °C in the samples containing xGnP-1.

The higher temperature peaks

were attributed to polymer regions entrapped between graphite nanoplatelets,

which retard the melting transition.

181

 

The melting behavior of PLA/xGnP-1 composites was further analyzed using
scanning probe thermal analysis in order to investigate possible connections
between agglomeration of xGnP and polymer molecular weight which were not
evidenced by the DSC analysis. For the neat PLA L209$ and PLA L21OS, the
deflection versus temperature curves were M-shaped: the tip deflection
increased upon heating from room temperature to approximately 80 °C,
decreased until the tip temperature reached approximately 100 °C, then
increased again until the onset of melting transition (Tm) (Figure 6.10).
Scanning probe thermal analysis revealed three major differences in the melting
behavior of the different molecular weight PLA and their composites with xGnP-1.

The recorded Tmo was lower (~ 150 °C) for PLA LZOQS than for PLA L21OS
(~159 °C). Lower temperatures were also recorded for the onset of melting
transition of PLA L2109$lenP-1 composites than for the PLA L21OS/xGnP-1
counterparts.

The deflection curves for PLA L209$l xGnP-1 composites were M-shaped,
similar to the curves recorded for the neat polymer (Figure 6.103). The deflection
versus temperature curve for PLA L21OS was also M-shaped (Figure 6.10b), but
linear increases of deflection with temperature were recorded for the majority of

the PLA L21OS samples containing xGnP-1.

182

 

Deflection (V)

% wt xGnP-1:

 

 

 

—0

 

 

80

Temperature (°C)

 

 

 

U"
V

Deflection (V)

 

A 0) co
l

 

 

 

 

F

20 40 60 80

l l

Tenperature (°C)

100 120 140 160 180 200 —8

—9

 

behavior.

Figure 6.10 Scanning probe thermal analysis deflection curves for a) PLA

L21OS/xGnP-1 and b) PLA L2OQS/xGnP-1 composites

the neat PLA L2098.

183

For all the PLA L209$l xGnP-1 composites, the deflections recorded were
lower than the deflection measured for the neat PLA L21OS (Figure 6.10a), while

higher deflections were recorded for all PLA L21OS/xGnP-1 composites than for

There appear to be three possible explanations for the differences in melting

First, differences in xGnP-1 dispersion in the two PLA matrices of

 

different molecular weight were considered to be responsible for the composites
behavior when subjected to nanoscale thermal analysis. Second, the extremely
high heating rate employed during scanning probe thermal analysis was
suspected to have enabled the identification of polymer chain re-arrangements
that were not detectable by conventional thermal analysis (DSC). Third,
differences in thermal conductivities of PLA L2098/xGnP-1 and PLA
L21OS/xGnP-1 were explored since tip deflection depends on the thermal and
electrical properties of the surface. This hypothesis is currently under
consideration.

The possibility of different dispersion of xGnP-1 in the two molecular weight
PLA samples was also indicated by thennomechanical analysis (DSC and DMA).
DMA results, in particular, indicated a smaller extent of improvement in storage
modulus in the case of composites prepared with the lower molecular weight
polymer. It is likely that xGnP-1 dispersed better in PLA L21OS, which was better
nucleated by the nanoplatelets, generating more polymer crystallites and fewer
regions in which xGnP-1 agglomerated. Dispersion is a key issue in processing
polymer nanocomposites, and has a strong impact on mechanical properties. As
opposed to phase separation — occurring when incorporating micro-size fillers in
polymers —, as the size of the fillers in the polymer matrices decreases,
dispersion and agglomeration become the important problems to be overcome
when attempting to obtain high-perfomiance polymer composites.18

In the PLA/xGnP-1 systems studied, more uniform crystalline structures

would lead to higher probabilities that the nano-TA tip encounters polymer

184

spherulites rather than regions where xGnP-1 agglomerates. The tip deflection
would be expected to be higher for polymer-rich regions.

Scanning probe thermal analysis is a surface technique, essentially different
from DSC, which is a bulk characterization method. Besides this important
difference, the heating rates used by the two techniques are significantly
different. The DSC runs investigating melting and crystallization were conducted
at heatingl cooling rates of 3 °Clmin, while during the nano-TA experiments the
tip was heated extremely fast (600 °Clmin). This difference might account for the
ability of scanning probe thermal analysis to detect polymer re-organization
phenomena on the sample surface which are not evident in larger volume
samples heated at slower rates during the DSC experiments.

In addition to crystallization from the melt, PLA undergoes cold crystallization,
upon heating from room temperature. For neat PLA samples used in this study,
cold crystallization occurred above the glass transition temperature (59-61 °C)
and below the melting transition temperature (179 °C) (Figure 6.11, and
Table 6.4).

Figure 6.11 and Table 6.4 also show that the extent of cold crystallization
depended on heating rate for both PLA LZOQS and PLA L21OS. For both
polymers, the temperatures of onset of cold crystallization (77°C), and the
temperatures of the cold crystallization peaks (ch) were higher for higher heating
rates. The peaks were larger and sharper for the lower molecular weight
polymer. Melting of PLA/xGnP—1 composites was also analyzed using DSC

(Figure 6.12), and did not indicate cold crystallization processes.

185

Cold crystallization of PLA in the presence of xGnP-1 was exclusively
detected by the nanoscale thermal analysis and is believed to be responsible for
the M-shape of the deflection versus temperature curves obtained using

scanning probe thermal analysis.

 

 

 

 

 

 

 

   

   
  

 

 

 

 

 

 

 

0.8 -
a)
15
E
X
3
C
‘6
0
I
.08 — Heating rate (°Clminute):
--3 ——5 —10
-1.2 -
Temperature (°C)
0.8 — b)
0.4 -
3
E 0 —
><
:: 2 100 140
C
‘6 -o.4 —
£
-0 8 Heating rate (°Clminute):
' m3 —5 —10
-1.2 <
Temperature (°C)

 

 

Figure 6.11 DSC therrnograms showing the cold crystallization and melting of
a) PLA LZOQS and b) PLA L21OS at three heating rates

186

 

 

Heat flux (W/g)

 

 

-0.6 ‘

 

 

C-
v

 

 

 

15
S
X
3
c
‘5
d.)
I
-0.5 — “’ 5 “6
—- 7 —8
-O.6 - '—9
Temperature (°C)

 

 

 

Figure 6.12 DSC therrnograms showing the melting of a) PLA L2098/xGnP-1
and b) PLA L21OS/xGnP-1 composites containing increasing
amounts of xGnP-1

187

Conclusions

Scanning probe thermal analysis of two different molecular weight poly(L-
lactic acid) samples and their composites with exfoliated graphite nanoplatelets
revealed essential differences in cold crystallization of the two polymer samples
in the presence of the nanoplatelets as well as the dispersion of xGnP in the
polymers. The difference in melting and crystallization behavior of the two
polymers was not detectable by differential scanning calorimetry, which by itself
could not explain the differences in the mechanical properties of the composites.

Nanoscale investigations of the thermal transitions occurring in polymer
nanocomposites were proven to be important for the characterization of the
dispersion of nanosize reinforcements, as well as for understanding the extent of
their nucleating effect on semicrystalline matrices. The results show that xGnP is

better dispersed in the higher molecular weight polylactic acid.

188

 

References

10.

Miloaga, D. G.; Hosein, H. A.; Rich, M. J.; Kjoller, K.; Drzal, L. T.,
Scanning Probe Thermal Analysis of Polylactic Acid/Exfoliated Graphite
Nanoplatelet (xGnPW) Nanocomposites. Journal of Biobased Materials
and Bioenergy 2008, 2, (1 ), 78.

Dinwiddie, R. B.; Pylkki, R. J.; West, P. E., Thermal conductivity contrast
imaging with a scanning thermal microscope. Thermal Conductivity 1994,
22, 668.

Balk, L. J.; Maywald, M.; Pylkki, R. J., Nanoscopic detection of the thermal
conductivity of compound semiconductor materials by enhanced scanning
thermal microscopy. Microscopy of Semiconducting Materials 1995, 1995;
Vol. 146, pp 655.

Hammiche, A.; Hourston, D. J.; Pollock, H. M.; Reading, M.; Song, M.,
Scanning thermal microscopy: Subsurface imaging, thermal mapping of
polymer blends, and localized calorimetry. Journal of Vacuum Science &
Technology 8 1996, 14, (2), 1486.

Hammiche, A.; Reading, M.; Pollock, H. M.; Song, M.; Hourston, D. J.,
Localized thermal analysis using a miniaturized resistive probe. Review of
Scientific Instruments 1 996, 67, (12), 4268.

Reading, M, Price, D.M., Grandy, D.B., Smith, R.M., Bozec, L., Conroy,
M., Hammiche, A., Pollock, HM, Micro-thermal analysis of polymers:
Current capabilities and future prospects, Macromolecular Symposia
2001, 167, 45.

Blaine, R. L.; Slough, C. 6.; Price, D. M. In Micro thermal analysis
calibration, repeatability and reproducibility, 27th North American Thermal
Analysis Society Annual Meeting and Exposition, Atlanta, GA, 1999;
Atlanta, GA. 1999.

Miloaga, D. (3.; Hosein, H. A. A.; Misra, M.; Drzal, L. T., Crystallization of
poly(3-hydroxybutyrate) by exfoliated graphite nanoplatelets. Journal of
Applied Polymer Science 2007, 106, (4), 2548.

Kalaitzidou K, F. H., Askeland P, Drzal L.T., The nucleating effect of
exfoliated graphite nanoplatelets and their influence on the crystal

structure and electrical conductivity of polypropylene nanocomposites.
Journal of Materials Science 2007 (Published online October 25, 2007).

Ray, S. 8.; Maiti, P.; Okamoto, M.; Yamada, K.; Ueda, K., New

polylactide/layered silicate nanocomposites. 1. Preparation,
characterization, and properties. Macromolecules 2002, 35, (8), 3104.

189

 

11.

12.

13.

14.

15.

16.

17.

18.

Ray, S. S.; Okamoto, M., Biodegradable polylactide and its
nanocomposites: Opening a new dimension for plastics and composites.
Macromolecular Rapid Communications 2003, 24, (14), 815.

Phang, I. Y.; Ma, J. H.; Shen, L.; Liu, T. X.; Zhang, W. D., Crystallization
and melting behavior of muIti-walled carbon nanotube-reinforced nylon-6
composites. Polymer International 2006, 55, (1 ), 71.

Maiti, P.; Nam, P. H.; Okamoto, M.; Hasegawa, N.; Usuki, A., Influence of
crystallization on intercalation, morphology, and mechanical properties of
polypropylene/clay nanocomposites. Macromolecules 2002, 35, (6), 2042.

Haggenmueller, R.; Fischer, J. E.; Winey, K. I., Single wall carbon
nanotube/polyethylene nanocomposites: Nucleating and templating
polyethylene crystallites. Macromolecules 2006, 39, (8), 2964.

Valentini, L.; Biagiotti, J.; Kenny, J. M.; Santucci, 8., Effects of single-
walled carbon nanotubes on the crystallization behavior of polypropylene.
Journal of Applied Polymer Science 2003, 87, (4), 708.

Homminga, D.; Goderis, B.; Dolbnya, l.; Groeninckx, G., Crystallization
behavior of polymer/montmorillonite nanocomposites. Part II. lntercalated
poly (epsiIon-caprolactone)/montmorillonite nanocomposites. Polymer
2006, 47, (5), 1620.

Homminga, D.; Goderis, B.; Dolbnya, I.; Reynaers, H.; Groeninckx, G.,
Crystallization behavior of polymer/montmorillonite nanocomposites. Part
I. lntercalated poly(ethylene oxide)lmontmorillonite nanocomposites.
Polymer 2005, 46, (25), 1 1359.

Ray, S. S.; Yamada, K.; Okamoto, M.; Ueda, K., Control of
biodegradability of polylactide via nanocomposite technology.
Macromolecular Materials and Engineering 2003, 288, (3), 203.

190

 

CHAPTER 7

CONCLUSIONS

Understanding how the crystallization of PHB and PLA is affected by addition
of exfoliated graphite nanoplatelets is essential for designing multifunctional
nanocomposites, since xGnP act as extremely efficient nucleating agents. The
nucleation and growth of the crystalline structures formed in the presence of
xGnP govern and also limit the properties of the nanocomposites. This research
was directed at understanding the interactions of xGnP with PHB and PLA, with a
goal that extended beyond filled systems, considering dimensional similarities
between physical characteristics of polymers and the nanosize reinforcing
particles, and exploring the polymer-particle interactions responsible for the
improved properties of the polymer nanocomposites. The intemal-out approach
presented here offered quantitative information about the amount of xGnP
required for nucleating purposes, and outlined the crystallization behavior of
PHB/xGnP and PLA/xGnP composites incorporating higher amounts of graphite
nanoplatelets in relation to other factors affecting the systems’ behavior.

Amounts of xGnP-1 up to 1 % wt suffice for nucleating purposes, triggering a
typical heterogeneous nucleating mechanism: in the presence of xGnP,

crystallization from the melt is thermodynamically favored, maximum attainable

191

crystallinity is achieved faster and at higher temperatures. A kinetic study of the
PHB/xGnP crystallization, employing a modified form of the Avrami equation,
revealed that the presence of xGnP increased the PHB crystallization
temperature, as well as the crystallization rates. Optical microscopy and atomic
force microscopy investigations revealed the reduction in the size of PHB and
PLA spherulites, indicating that the xGnP-nucleated polymers form smaller and
more irregular crystalline structures and spherulitic morphologies specific for the
neat polymers are no longer recognizable in the composites. X-ray diffraction
showed that both PHB/xGnP and PLA/xGnP crystallize in the same crystal
structure as the neat polymers.

Although microscopy techniques did not indicate significant differences
between different composites containing amounts of xGnP higher than 3 % wt,
for which properties were shown to depend on processing conditions, and on
polymer molecular weight, differential scanning calorimetry revealed qualitative
differences in crystallinity occurring in PLA/xGnP composites at higher xGnP
loadings. DSC analyses successfully illustrated the competition between two
effects of xGnP in polylactides matrices: nucleation, and retardation of crystal
growth.

As part of this research, thermal analysis was further refined using scanning
probe thermal analysis (nano-TA) as a novel characterization technique for
investigating the morphology of these polymer nanocomposites. Nanoscale
investigations of the thermal transitions in PLA/xGnP nanocomposites were

important for the characterization of the dispersion of nanosized reinforcements,

192

as well as for understanding the extent of their nucleating effect on
semicrystalline matrices. The information obtained using this technique
complemented the conclusions derived from classical thermal analyses and
offered additional insight on the relation between polymer confinement and
molecular weight. The results are particularly important since they may be
applied to nanocomposites utilizing similar semicrystalline polymer matrices.

Nanocomposite properties of the PLA/xGnP systems were investigated with a
special focus on the relation between the nucleating effect of xGnP and change
in mechanical properties and thermal conductivity. Although several processing
methods were evaluated, the classical solvent mixing approach lead to highest
improvement in all the properties evaluated. Using solvent mixing and
compression molding, 30 % improvement PLA modulus was achieved with
incorporation of as low as 0.01 % wt xGnP-1.

Regardless of the mixing and processing approach used, polymer
confinement effects were observed in PLA/xGnP composites containing higher
xGnP content. The addition of xGnP in excess of that required for nucleation
resulted in the agglomeration of xGnP, and retardation of crystal growth.

Nanocomposites prepared both by melt mixing and solvent mixing showed
that the same amount of xGnP was required for achieving electrical and the
thermal conductivity. The nucleating effect of xGnP is a limiting factor in
achieving high electrical and thermal conductivity in PLA/xGnP nanocomposites
since the formation of large numbers of crystalline structures, thermodynamically

favored by xGnP, isolates xGnP and prevents the direct contact of graphite

193

 

nanoplatelets in the resulted nanocomposites. The results suggest the possibility
that not only the electron transport, but also the heat transport occurs through a
percolation mechanism.

Thermal conductivity of PLA/xGnP nanocomposites can be improved by
introducing a second phase to control the positioning of xGnP particles.
Polystyrene, an amorphous polymer immiscible with PLA, was selected as a
model polymer that when added to PLA and xGnP caused concentration of xGnP
at the polymer-polymer interface creating a continuous xGnP path. Future work
following this approach is suggested.

The findings are important for the design of nanocomposites from xGnP and
PHB, and PLA, since the results outline the consequences of adding increasing
amounts of graphite nanoplatelets to the polymers and approaches that can be
used to target various properties.

On a larger scale, the study is of general interest for nanocomposites from
similar semicrystalline polymer matrices in which nanoparticles act as nucleating

agents.

Future work

The recent results regarding thermal and electrical conductivity imparted by
xGnP to PLA matrices indicate that for achieving high conductivities in PLA and
similar semicrystalline polymers, the focus needs to be on desirable changes in

morphology that lead to lower percolation points.

194

 

Several ideas are suggested for achieving higher levels of thermal
conductivity in PLA nanocomposites with xGnP. Taking advantage of polymer
phase separation in binary blends is one possible approach that could prevent
the encapsulation of xGnP in crystalline structures nucleated in the polymer
matrix. The corresponding changes in mechanical properties are to be carefully
monitored if following this approach. We also suggest the investigation of
PLA/xGnP systems containing plasticizers, and the exploration of amorphous

polylactides, and PLA mixtures containing D and L forms.

195

 

   

l1111111111111111111