:1
. ,1 . 1214.:-
.. 5.

A... .
Juana. . I.

(.3513. .9.
lit: 1.!
23:35:... at
I». .3301
a if: 14.!

. .21....
‘- '0’

u :4. ‘2

«333:: \-

Ilitrnzu
(1‘99

 

a.
n:
0....

1.
a...“
. l.

I... . .
1:3.l1s‘1xubhnca

r56

:5\ i!!!

 

s...
Iailil... 1's.
...I(7LI ‘ ,e (y
.1: i?

 

 

 

 

'3:~.CPIL:.DL Lint}. Bk.

2001

This is to certify that the
dissertation entitled

CONTINUOUS MICROCELLULAR FOAMING OF
POLYVINYL CHLORIDE AND COMPATIBILIZATION OF
. POLYVINYL CHLORIDE AND POLYLACTIDE COMPOSITES

presented by

BHAVESH SHAH

has been accepted towards fulfillment
of the requirements for the

 

Doctoral degree in Forestry

 

 

MajorT’Fofessor’ 5 Signature
f///' 7 9/75”?“-

Date

MSU is an Affinnative Action/Equal Opportunity Institution

 

 

LIBRARY
Michigan State
University

 

 

 

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

Aug a {#0389

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/ClRC/DaleDue.indd-p.1

CONTINUOUS MICROCELLULAR FOAMING OF POLYVINYL CHLORIDE AND
COMPATIBILIZATION OF POLYVINYL CHLORIDE AND POLYLACTIDE
COMPOSITES
By

Bhavesh Shah

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Forestry

2007

ABSTRACT
CONTINUOUS MICROCELLULAR FOAMING OF POLYVINYL CHLORIDE
AND COMPATIBILIZATION 0F POLYVINYL CHLORIDE AND
POLYLACTIDE COMPOSITES
By
Bhavesh Shah

This dissertation focuses on overcoming existing limitations of WPCS which
prevent them from realizing their full market potential. These limitations include: i) lack
of a continuous extrusion process for microcellular foaming of polyvinyl chloride (PVC)
and its composites using supercritical fluids to reduce the high density of the WPCS, ii)
need for an efficient coupling agent for WPCS to overcome the poor compatibility
between wood and plastic, and iii) unproven use Of wood as a filler for the biopolymer
polylactide (PLA) to make “green” composites. These limitations were addressed
through experimentation to develop a continuous extrusion process for microcellular
foaming, and through surface modification of wood flour using natural coupling agents.

The effects of wood flour, acrylic modifier and plasticizer content on the
rheological properties of PVC based WPCS were studied using an extrusion capillary
rheometer and a two-level factorial design. Wood flour content and acrylic modifier
content were the major factors affecting the die swell ratio. Addition of plasticizer
decreased the true viscosity of unfilled and filled PVC, irrespective of the acrylic
modifier content. However, the addition of acrylic modifier significantly increased the
viscosity of unfilled PVC but decreased the composite viscosity.

Results of the rheological study were used to set baseline conditions for the
continuous extrusion foaming of PVC WPCS using supercritical C02. Effects of material

composition and processing conditions on the morphology of foamed samples were

investigated. Foamed samples were produced using various material compositions and
processing conditions, but steady-state conditions could not be Obtained for PVC. Thus
the relationships could not be determined.

Incompatibility between wood flour and PVC was the focus of another study.
The natural polymers chitin and chitosan were used as novel coupling agents to improve
interfacial adhesion between the polymer matrix and wood fiber. Results indicated that
addition of chitin and chitosan significantly increased the flexural properties and storage
modulus Of PVC WPCS, compared to composites without coupling agent. Significant
improvements were attained with 0.5 wt. % chitosan and with 6.67 wt. % chitin.

Based on the efficiency of chitosan as a coupling agent for PVC based WPCS, a
biodegradable composite using polylactide (PLA) and chitosan was developed. Wood
flour (0 — 40 wt. %) was evaluated as a filler for PLA composites and its effect on
mechanical, thermal and chemical properties was studied with and without chitosan (O —
10 wt. %). Addition of wood flour significantly increased the flexural and storage moduli
of PLA-wood flour composites, but had no effect on glass transition temperature (Tg).
Chitosan had no significant effect on any of the properties of the composites studied.

Development of an efficient and effective coupling agent for PVC wood
composite is a significant development which will increase performance while reducing
cost. Wood filled PLA composites can further expand WPCS into applications such as
packaging and automotive. Results from these studies have broadened the current
knowledge base for WPC products and will be useful in the continued expansion of wood

composites technology into a variety Of industries.

Dedicated to my Family and Karana, without them this would not have been

possible

iv

ACKNOWLEDGEMENTS

I would like to thank my major professor, Dr. Walters. Without him this thesis
would not have been completed. His continuous encouragement, guidance and
thoughtful insight helped me achieve this milestone. His moral and monetary support
kept me going through some of my most challenging times. I would like to also thank
Dr. Selke for serving as the dissertation advisor on my committee. Her invaluable
suggestions, guidance and patience helped me take my work from ideas to the finished
product. She was the driving force behind me. I appreciate all of the time she Spent
reading my dissertation and giving me constructive comments.

I am grateful to Dr. Drzal for giving me permission to use equipment at the
CMSC to finish this research and also for his positive and productive comments
throughout this dissertation. Without his support and help, I would not have been able to
finish the experimentation for this research. Dr. Bix helped and supported me throughout
my dissertation, and was always available with constructive comments and advice. She
encouraged me to stay positive when things were difficult. I am also thankful to Dr.
Heiden from Michigan Technological University for carefully reading this dissertation
and offering suggestions, editing help and guidance which made the finished product
better.

I would like to acknowledge the financial support Of the various funding agencies
and departments that supported parts of this work, including the USDA-CSREES Grant —

Advanced Technology Applications to Eastern Hardwood Utilization, College of

Agriculture and Natural Resources, MSU Graduate School and the Office of International
Students and Scholars.

I am thankful to all of my current and former co-workers, researchers and friends
from MUS including Jesse, Natalie, Justin, Jae-Woo, Jeanette, Li, Na, Katie, Sedric,
Smith and countless others for making my stay and work here memorable. I would
especially like to thank Karana for always being with me through all of the ups and
downs of my PhD and life.

I am grateful to the Faculty and Staff of the Department Of Forestry, and
especially Juli for her constant help and support throughout my program. We had some
wonderful times and She helped me to get all Of my paperwork done.

So many people have helped me out in big and small ways throughout this
process. I am sure I have forgotten to thank some of them by name, but know that I do
appreciate everyone who made my life a little easier during the time spent at MSU.

Finally, I would like to thank my family for their support and patience.

vi

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................................... x
LIST OF FIGURES ....................................................................................................... xi
LIST OF ABBREVIATIONS ....................................................................................... xiii
CHAPTER 1
OVERVIEW ................................................................................................................... 01
References ................................................................................................................... 06
CHAPTER 2
BACKGROUND, RATIONALE AND SPECIFIC HYPOTHESES ........................ O7
Foamed PVC/WOOd-Flour Composites ...................................................................... 07
Microcellular F oamea' Polymer ............................................................................. 08
Coupling Agents for PVC/WOOd-Flour Composites .................................................. 16
Polylactide ................................................................................................................... 2 1
Summary ..................................................................................................................... 24
Hypotheses .................................................................................................................. 26
References ................................................................................................................... 27
CHAPTER 3
ONLINE MEASUREMENT OF RHEOLOGICAL PROPERTIES OF
PVC/WOOD-FLOUR COMPOSITES ........................................................................ 33
Abstract ....................................................................................................................... 34
Introduction ................................................................................................................. 35
Experimental ............................................................................................................... 39
Materials ................................................................................................................ 39
Experimental Design and Compounding ............................................................... 41
Rheology Experiments ........................................................................................... 43
Results and Discussion ............................................................................................... 46
Statistical Analysis of the Die Swell Ratio ............................................................. 46
Statistical Analysis of the True Viscosity ............................................................... 52
Conclusions ................................................................................................................. 60

vii

References ................................................................................................................... 6 1

CHAPTER 4
CONTINUOUS EXTRUSION OF MICROCELLULAR F OAMED PVC AND ITS
WOOD FLOUR COMPOSITES USING SUPERCRITICAL CO; .......................... 64
Abstract ....................................................................................................................... 65
Introduction ................................................................................................................. 66
Experimental ............................................................................................................... 70
Materials ................................................................................................................ 70
Continuous foaming process .................................................................................. 70
Thermodynamics of Foaming ................................................................................ 73
Physical Characterization of Foams ..................................................................... 75
SEM Analysis .......................................................................................................... 75
Results and Discussion ............................................................................................... 77
Experimental Setup ................................................................................................ 77
Microcellular foaming of P V C ............................................................................... 78
Conclusions ...................................................................................................................... 87
References ........................................................................................................................ 88
CHAPTER 5
NOVEL COUPLING AGENTS FOR PVC/WOOD-FLOUR COMPOSITES ....... 91
Abstract ....................................................................................................................... 92
Introduction ................................................................................................................. 93
Literature Review on Chitin and Chitosan ............................................................ 96
Experimental ............................................................................................................... 100
Materials ................................................................................................................ 100
Composite Manufacturing ..................................................................................... 102
Property Testing ..................................................................................................... 102
Results and Discussions .............................................................................................. 104
F lexural Properties ................................................................................................ l 04
Dynamic Mechanical Properties ........................................................................... 108
Conclusions ................................................................................................................. 1 12
References ................................................................................................................... 1 13

viii

CHAPTER 6
EFFECTS OF WOOD FLOUR AND CHITOSAN ON MECHANICAL,

CHEMICAL AND THERMAL PROPERTIES OF POLYLACTIDE .................... 115
Abstract ....................................................................................................................... 116
Introduction ................................................................................................................. 1 17
Experimentation .......................................................................................................... 1 23

Materials ................................................................................................................ 123
PLA- Wood Flour Composite Manufacturing ......................................................... 123
Mechanical Property Testing ................................................................................. 125
Statistical Analysis ................................................................................................. 125
Dynamic Mechanical Analysis (DMA) .................................................................. 125
Surface Analysis ..................................................................................................... 126
Scanning Electron Microscopy (SEM) ................................................................... 126
Results and Discussions ............................................................................................ 127
Mechanical Properties ........................................................................................... 127

F T IR Analysis ......................................................................................................... 132
Dynamic Mechanical Analysis ............................................................................... 136
Morphology ............................................................................................................ 140
Conclusions ................................................................................................................. 144
References ..................................................................................................................... 145

CHAPTER 7

SUMMARY AND RECOMMENDATIONS ............................................................... 148

APPENDIX A ................................................................................................................. 153

APPENDIX B ................................................................................................................. 149

ix

Table 3.1

Table 3.2

Table 3.3

Table 3.4

Table 4.1.

Table 4.2.

Table 5.1.

Table 5.2.

Table 6.1.

Table 6.2.

Table A. 1.

Table B. 1.

Table B. 2.

LIST OF TABLES

Formulations Used in Neat PVC and PVC/WOOd-Flour Composites ....... 40

Experimental Design Matrix in Terms of Actual and Coded Factor Levels
Generated by Desi gn-Expert® Software ................................................... 42

Probability and Summary Statistics for Two-Level Factorial Model ........ 48
Effect of Acrylic Modifier on the Storage Modulus (Elastic Modulus) and
Loss Modulus (Viscous Modulus) of Neat PVC and PVC/Wood-Flour'
Composites ................................................................................................. 58

Formulations Used in Manufacturing Microcellular Foamed PVC and its
Wood-Flour Composites ............................................................................ 71

Different Processing Conditions used to Manufacture Microcellular

F oamed PVC and Its Composites .............................................................. 82
Formulations Used in PVC/WOOd-Flour Composites ............................. 101
Effect of Coupling Agent on the Dynamic Mechanical Properties of Neat

Rigid PVC and PVC/WOOd-Flour Composites ........................................ 111
Individual Batch Composition ................................................................. 124

FTIR Band Assignment of Major Peaks in PLA and Wood Flour .......... 133
Chemical Composition Of Wood ............................................................. 153

Mechanical Properties of the PVC/wood-flour Composites made with
different concentration of coupling agent compared with the control
sample ...................................................................................................... 166

Mechanical Properties of the PLA/wood-flour Composites made with
different concentration Of coupling agent ................................................ 167

Figure 2.1
Figure 2.2

Figure 3.1

Figure 3.2

Figure 3.3

Figure 3.4

Figure 4.1.
Figure 4.2.
Figure 4.3.

Figure 4.4.

Figure 5.1.

Figure 5.2.

Figure 5.3.

Figure 5.4.

LIST OF FIGURES

Microcellular Foaming Process ................................................................. 10
Schematic of a Batch Foaming Process ..................................................... l 1

Perturbation plot of die swell against three investigated variables. A
represents the wood flour content; B is the acrylic modifier content; and C
is the plasticizer content ............................................................................. 50

Die swell ratio as a function of the interaction between wood flour content
and at O and 8 phr acrylic modifier (K-400) content. ................................ 51

True viscosity at different shear rates: (a,b) at 500 s'1 (c,d) at 1200 s'1 and
(e,t) at 2000 SI as a function Of the interaction between wood flour and at
0 and 4 phr plasticizer content: (a,c,e) no acrylic modifier and (b,d,t) 8 phr

acrylic modifier .......................................................................................... 55
Entanglements of PVC Chains and Acrylic Modifier ................................ 59
Schematic of Continuous Microcellular Foaming Process. ....................... 72
SEM Micrographs of HDPE at (a) 100x and (b) 10,000X ........................ 80

SEM Micrographs Of PVC foamed at 3 wt. % CO; (a) 20X and (b) 50X .84

SEM Micrographs of PVC foamed at (a) 2 wt. % C02 and (b) 1 wt. %

C02 ............................................................................................................. 86
y—Amino Propyltriethoxy Silane Reaction Mechanism with Wood and

PVC ............................................................................................................ 95
Structure of (a) cellulose, (b) chitin and (c) chitosan ................................. 97

F lexural strength of PVC and its composites with varying concentration of
coupling agent: (a) chitin and (b) chitosan. Different letters above the bars
indicate significant differences between the uncoupled composite and

composites with chitin or chitosan at P < 0.05 ........................................ 105

F lexural modulus of PVC and its composites with varying concentration
of coupling agent: (a) chitin and (b) chitosan. Different letters above the
bars indicate significant differences between the uncoupled composite and
composites with chitin or chitosan at P < 0.05 ........................................ 106

xi

Figure 5.5.

Figure 6.1.

Figure 6.2.

Figure 6.3.

Figure 6.4.

Figure 6.5.

Figure 6.6.

Figure 6.7.

Figure 6.8.

Figure 6.9.

Figure 6.10.

Figure A.l.

Figure A.2.

LIST OF FIGURES (CONT’D)

DMA curve of rigid PVC/wood—flour composite without coupling

agent ......................................................................................................... 110
Structures of (a) Polylactide (b) Poly(e-caprolactone) (c) Cellulose and (b)
Chitosan ................................................................................................... 120
Possible Chemical Reaction between Amine and Ester Group ............... 121

F lexural Properties of PLA and its Composites with Varying
Concentration of Wood Flour (a) Flexural Strength and (b) F lexural
Modulus. Different letters above the bars indicate statistically significant
differences at P < 0.05 ............................................................................. 129

Predicted and Experimental Values of F lexural Modulus with Increasing
Volume Of Fiber ....................................................................................... 130

F lexural Properties of PLA/Wood Flour Composites with 40 wt. % Wood
Flour and Varying Concentration of Chitosan (a) Flexural Strength and (b)
Flexural Modulus. Different letters above the bars indicate statistically
Significant differences at P < 0.05 ........................................................... 131

FTIR Spectra of (a) PLA, (b) Maple Wood Flour, (c) PLA Composite with
40 wt. % Wood Flour and (d) PLA Composite with 40 wt. % Wood Flour
and 10 wt. % Chitosan ............................................................................. 134

Temperature Dependence of (1) Storage Modulus and Tan Delta and (11)
Loss Modulus of (a) PLA, (b) PLA Composite with 20 wt. % Wood Flour
and (c) PLA Composite with 40 wt. % Wood Flour ................................ 137

Temperature Dependence of (1) Storage Modulus and Tan Delta and (11)
Loss Modulus Of PLA composite with 40 wt. % Wood Flour and Varying

Amount of Chitosan (a) 0%, (b) 2.5% and (c) 5% ................................... 139
SEM Images of PLA Composites with 20 wt. % Wood Flour at Different
Magnifications: (a, c) 100X and (b, d) 400x ........................................... 141
SEM Images of PLA Composite with 40 wt. % Wood Flour at Different
Magnifications: (a, b) 100X and (c, d) 400x ........................................... 143
Structure of PVC (where n = 700 — 1500) ............................................... 156
Polymerization of Polylactide .................................................................. 159

xii

LIST OF ABBREVIATIONS

ABS -- Acrylonitrile Butadiene Styrene
ANOVA -— Analysis Of Variance
CPE -- Chlorinated Polyethylene
DEHP —— Di-2-ethy1hexyl phthalate
DMA -- Dynamic Mechanical Analysis
F TIR -- Fourier Transform Infrared Spectroscopy
HDPE -— High Density Polyethylene
HIPS -- High Impact Polystyrene
MF I -- Melt Flow Index
MW -- Molecular Weight
PC -- Polycarbonate
PCL -- Polycaprolactone
PE -- Polyethylene
PLA -- Polylactide
PP -- Polypropylene
PS -- Polystyrene
PVC -- Polyvinyl Chloride
SEM -- Scanning Electron Microscopy
Tg -- Glass Transition Temperature

WPCS -— Wood-Plastic Composites

xiii

CHAPTER 1

OVERVIEW

Wood—plastic composites (WPCS) are an emerging class of engineering materials
which combine the benefits of both wood and plastic. The market for WPCS in North
America was around 1.2 billion pounds in 2004 and is expected to reach 2 billion pounds
by the year 2009, at an average annual growth rate of 10%. Major applications of WPCs
include decking, railing, fencing, window and door frames and various automotive parts
[1]. WPCS use wood flour or fiber as a filler or reinforcing agent in a plastic matrix.
Wood has several advantages over conventional mineral fillers (talc, calcium carbonate,
clay) such as low density, low cost and it is less abrasive to processing equipment. There
are several other natural cellulosic fillers used in manufacturing composites such as flax,
hemp, kenaf, newsprint fiber, jute, rice husk, sisal, coir, bamboo, etc. Natural fiber based
fillers are abundantly available, recyclable, and biodegradable which makes them the best
alternative to mineral fillers. Of all the natural cellulosic fillers, wood flour is the most
widely used filler in manufacturing composites in the United States.

WPCS can be made with various different plastics such as polyethylenes (PE),
polypropylenes (PP), polyvinyl chloride (PVC), polystyrene (PS) and several others.
PVC is the second most widely used polymer for manufacturing WPCS though the
market is dominated by PE. PVC as a matrix has several advantages over polyolefins
(PE, PP, etc.) such as high stiffness, good weatherability, ease of customizing the final

product properties by changes in formulation, and good paintability of the final product.

Wood flour is used extensively as a filler and Offers several advantages, but it has
limited thermal stability and high moisture absorption; further, the addition of wood fiber
or flour into the plastic reduces its strength properties (tensile and flexural strength)
compared to the unfilled plastic. This is mainly due to the poor dispersion of fiber and
lack of strong interaction between the hydrophilic wood fibers and hydrophobic polymer
matrix [2]. Coupling or compatibilizing agents can increase the dispersion of fiber in the
matrix and form polymer—fiber bonds by modifying the wood surface [3]. The use of
coupling agents in manufacturing WPCS has been known for several years. However,
there has been limited success with coupling agents for PVC/wood-flour composites.
Thus, an effective coupling agent for PVC/wood-flour composites is still needed.

Although WPCs have been commercialized, their potential for use in many
industrial applications has been limited because of their brittleness, lower impact resistance,
and higher density compared to virgin plastics and/or solid wood. The high density of
WPCS, which is almost twice that of solid wood, makes them heavier for handling and
end use applications which may also hinder their growth in various applications. These
drawbacks have been addressed by production of microcellular-foamed structures in
WPCS which reduce the density of the composites, thus saving the cost of material while
increasing mechanical properties such as impact strength [4-7]. Most of the research on
microcellular foaming of PVC and its composites has been done using a batch foaming
process. However, the batch foaming process used to generate cellular foamed structures
. is not likely to be implemented in the industrial production of foams because it is time
consuming and not cost-effective. To manufacture microcellular foamed PVC product

economically and efficiently a continuous extrusion process using supercritical fluids

needs to be developed. This likely has not been done due to the complex processability
and intricacies of PVC formulation. Each component in the formulation has a different
effect on the behavior of PVC melt which controls the foaming process and these
interactions need to be understood before actual foaming process can be successfully
implemented.

Increasing use of PVC in manufacturing WPCS has led to a growing concern
about its recyclability. Recycling of PVC is a complicated process, due to the presence
of several different additives in the product being recycled. Additionally, the type and
amount of additives used varies from product to product. It is difficult to obtain desired
properties in products made from recycled PVC as it contains an unknown mix of
additives. PVC is also prone to thermal degradation during the recycling process, which
may cause it to break down and release hydrochloric acid. Due to these recycling issues
some manufacturers are replacing PVC in various applications with other commodity
polymers.

Consumer awareness about recycling and environmental impact of synthetic
polymers in combination with government initiatives has lead to growth of the
biodegradable polymer industry in recent years. Polylactide (PLA) is one of the most
widely used and produced biobased polymers. The properties of PLA are comparable to
most commodity polymers like PP, PE, PVC, and PS. Because of this PLA is a good
alternative to synthetic polymers in manufacturing natural fiber composites. Several
researchers have shown PLA can be successfully used as a matrix for natural fiber
composites [8-10]. However, there is no published literature on using wood as a filler in

PLA composites.

Even though PLA natural fiber composites are manufactured, poor dispersion and
weak interfacial bonding of the hydrophilic fibers with the hydrophobic polymer matrix
leads to poorer strength properties than unfilled polymer [9-11]. Maleic anhydride and
various maleated compounds have been used to compatibilize the components in PLA
composites with limited success. An effective coupling agent for PLA natural fiber
composite is still needed to improve the strength properties. A natural material that
would function as a coupler would be desirable for biobased composites as it would be
completely renewable, resulting in a “green” composite.

WPCS have enormous potential to replace conventional materials in a variety of
applications, but there are major limitations that need to be overcome before this
possibility can be realized. These limitations include, i) lack of a continuous extrusion
process for microcellular foaming of PVC and its composites using supercritical fluids,
ii) need for an efficient coupling agent for PVC wood composites that would overcome
the poor compatibility between wood and plastic, and iii) unproven use of wood as a filler
for biopolymer PLA and need for better interfacial adhesion of PLA natural fiber
composites.

This research aims to advance wood composites technology and utilization via
experimentation focused on improving WPCS properties and broadening the applications

for WPCS. This work has three main areas of focus:

1. To develop a continuous extrusion process for foaming PVC and its wood flour
composites using supercritical fluids, which will greatly increase the ease of

producing microcellular foamed products in a way that is efficient and

economical. This will result in improved end-product quality and reduced

material consumption compared to unfoamed products.

. TO develop a new class Of coupling agents for PVC based WPCS. This will
improve the interfacial adhesion between polymer and filler and result in stronger

performing composite products.

. To demonstrate the use of wood as a filler for PLA, and to improve the properties
of these composite with a natural coupling agent. This will result in
environmentally friendly, renewable composite products, which are competitive

with those made from synthetic polymers, yet can be recycled.

10.

11.

REFERENCES

Clemons, C.M., "Wood Plastic Composite in the United States: The Interfacing of
Two Industries", Forest Products Journal, 52(6): 10-18 (2002).

Matuana, L.M., Woodhams, R.T., Balatinecz, J.J., and Park, C.B., "Influence of
interfacial interactions on the properties of PVC/cellulosic fiber composites",
Polymer Composites, 19(4): 446-455 (1998).

Li, Q. and Matuana, L.M., "Surface Of cellulosic materials modified with

functionalized polyethylene coupling agents", Journal of Applied Polymer
Science, 88(2): 278-286 (2003).

Matuana, L.M., Park, OH, and Balatinecz, J .J ., "Cell morphology and property
relationships of microcellular foamed PVC/wood-fiber composites", Polymer
Engineering and Science, 38(1 1): 1862-1872 (1998).

Matuana, L.M., Park, CB, and Balatinecz, J.J., "Processing and cell morphology
relationships for microcellular foamed PVC/wood-fiber composites", Polymer
Engineering and Science, 37(7): 1137-] 147 (1997).

Matuana-Malanda, L., Park, CB, and Balatinecz, J.J., "Characterization of
microcellular foamed PVC/cellulosic-fiber composites", Journal of Cellular
Plastics, 32(5): 449-469 (1996).

Mengeloglu, F. and Matuana, L.M., "Foaming of rigid PVC/wood flour
composites through a continuous extrusion process", Journal of Vinyl & Additive
Technology, 7(3): 142-148 (2001).

Oksman, K., Skrifvars, M., and Selin, J.F., "Natural fibres as reinforcement in
polylactic acid (PLA) composites", Composites Science and Technology, 63(9):
1317-1324 (2003).

Lee, S.-H., Ohkita, T., and Kitagawa, K., "Eco-composite from poly(lactic acid)
and bamboo fiber", Holzforschung, 58(5): 529-536 (2004).

Huda, M.S., Mohanty, A.K., Drzal, L.T., Schut, E., and Misra, M., ""Green"
composites from recycled cellulose and poly(lactic acid): Physico-mechanical and
morphological properties evaluation", Journal of Materials Science, 40(16):
4221-4229 (2005).

Huda, M.S., Drzal, L.T., Misra, M., Mohanty, A.K., Williams, K., and Mielewski,
DE, "A Study on Biocomposites from Recycled Newspaper Fiber and Poly(lactic
acid)", Industrial & Engineering Chemistry Research, 44(15): 5593-5601 (2005).

CHAPTER 2

BACKGROUND, RATIONALE AND SPECIFIC HYPOTHESES

This section will provide background material and justification for the research
work conducted in this project. The current state of the knowledge on PVC and PLA
wood composites will be discussed, along with the issues that currently limit their use in
various applications. Specific hypotheses will be developed to address ways to overcome

these limitations.

F oamed PVC/Wood-Flour Composites:

WPCS use wood flour or fiber as a filler or reinforcing agent in a plastic matrix
(see Appendix A for information on wood properties). WPCS are increasingly gaining
popularity in diverse applications, some of which are replacements for conventional
materials. PVC wood composites offer several advantages over polyolefin based wood
composites such as good weatherability, high stiffness, ease of customizing the final
product properties by changes in formulation, and good paintability of the final product.
However, PVC/wood-flour composites have a major disadvantage of high density due to
the presence of PVC (see Appendix B for information on PVC properties). The high
specific gravity of PVC compared to wood leads to an increase in the density of the final
product, making it heavier for handling and end use applications. Addition of wood to
PVC also significantly reduces the impact strength of the PVC matrix. These drawbacks
have been addressed by production of microcellular-foamed structures in WPCS which

reduce the density of the composites, thus saving the cost of material and increasing

mechanical properties such as impact strength [1-4]. Foaming also improves the
fastening properties of the composites. Matuana et a1. [3] showed that introducing a
microcellular structure into PVC/wood-flour composites can increase its impact strength
while reducing the overall density of the composites. Another approach used to
overcome the drawbacks of WPCS is to use impact modifiers in the formulations.
However, impact modification of wood/plastic composites does not enhance their
ductility or reduce the density of the products [5]. When similar impact modifiers were
used in the microcellular foamed formulations, they had a deleterious effect on the cell
structure [6]. Addition of impact modifier increases the diffusion coefficient of gas in the
samples, resulting in lower density reduction (void fraction) in the foamed sample. The
increased gas diffusion coefficient increases the gas loss rate which reduces the amount
of gas available for the foaming process. Therefore, the use of impact modifiers in the

foam formulations should be avoided.

Microcellular F 0amed Polymer:

Microcellular polymers are foamed plastics characterized by average cell size less
than 10 um and cell-population density greater than 109 cells/cm3. Manufacturing
microcellular foam involves three basic steps viz., i) saturation ii) nucleation and iii) cell
growth and stabilization. These basic steps are illustrated in Figure 2.1. Microcellular
foam can be produced by batch, semi-continuous and continuous processes. Irrespective
of the process, a blowing agent is required to generate the foamed structure in the
polymer. Two major classes of blowing agents are used in the foaming process, chemical

and physical blowing agents. Chemical blowing agents work by breaking down the

compound into a gas at a specific temperature during processing. In contrast, physical
blowing agents are typically inert gases injected into the polymer at a certain stage of
processing. Since they are generally more eco-friendly, physical blowing agents are
usually preferred over chemical blowing agents which may generate environmentally
undesirable products. C02 and N2 are the most common physical blowing agents being
used in manufacturing microcellular foaming, but the most favored one is C02. This is
due to higher diffusivity and solubility of C02 in most polymers compared to N2.

Most of the early research on microcellular foaming involved a batch process. A
typical setup of a batch process is shown in Figure 2.2. In the typical batch foaming
process, a pro-prepared solid sample is saturated with a physical blowing agent under
specific conditions of temperature and pressure in a sealed chamber (Figure 2.2a). The
saturation process can last up to several days depending on the saturation temperature,
pressure, sample geometry and polymer system. Once the sample is saturated with the
gas it is transferred to a heated oil bath, the temperature of which is maintained above the
glass transition temperature (T3) of the polymer being foamed (Figure 2.2b). The
increase in the temperature drives the nucleation of bubbles in the sample, and the time
the sample is kept at that temperature determines the growth of the bubbles. Several
amorphous, semi-crystalline and crystalline polymers such as HDPE [7], PP [8], PVC

[9], PS [10], PC [11], etc. have been foamed using this technique.

 

0000000

0
0
0
0
O
0
0
0

90

0
9
O
O
0
9
O
0

 

Microcellular Foam

Figure 2.1 Microcellular Foaming Process

    

Nucleation

presswized cavity glycerin bath

           

pressure vessel o-ring

control

(a) Gas saturation process (b) Foaming process

inert gas

Figure 2.2 Schematic of a Batch Foaming Process [12]

In spite of the ease and simplicity of batch foaming, it cannot be used for
industrial applications due to the time consuming process of sample saturation, which
makes the whole process uneconomical. To overcome these drawbacks, a continuous
process of manufacturing microcellular foam using an extrusion technique was proposed
by several researchers [13-16]. The continuous extrusion process has been successfully
used for several polymers, but to date there has been no available literature on using this
process for PVC and its wood flour composites with online injection of supercritical fluid
as the blowing agent. The research carried out on continuous extrusion process of
microcellular foaming of PVC and its composites has been mainly focused on chemical
blowing agents, although the use of moisture as a foaming agent has also been reported in
the literature [17]. Chemical blowing agents require extra care during dry blending to
avoid migration, which results in density variations in the final product. Apart from this,
chemical blowing agents have several other disadvantages including difficulties in
adjusting foam densities during processing [18]. Physical blowing agents such as carbon
dioxide (CO2), nitrogen (N2) and argon (Ar) are economical and eco-friendly, which
makes them a better alternative to chemical blowing agents. One of the goals of this
research is to manufacture microcellular foamed PVC and its composites in a continuous
extrusion process using supercritical fluids.

Before continuous microcellular foaming can be implemented for PVC and its
composites, the effects of material compositions and extrusion processing conditions on
the melt viscosity of the polymer must be investigated. The melt viscosity of the polymer
matrix has a major effect on the processability of the polymer and the end product

quality. Rheological data can be a vital part of process design and setting correct

12

processing conditions for the polymer. On the basis of rheological data, process variables
can be modified to avoid processing problems such as melt fracture, sharkskin, and die
swell. The viscosity of the polymer melt also plays a vital role in cell nucleation and
growth during the foaming process. Therefore, it is of great importance to study the
effect of different additives and processing conditions on the rheology of the polymer.

The rheological properties of PVC have been continuously studied for several
decades and are still of interest to researchers [19-23]. Collins and Krier [19] have found
two different activation energies for PVC. A low activation energy for viscosity was
observed for a low temperature region characterized by particulate flow, while a high
activation energy was found for the high temperature region characterized by molecular
flow. The temperature at which the transition from the low to the high temperature
region occurred was a function of shear stress. The effect of low molecular weight
polymers like acrylate, polypropylene terephthalate and PVC polymer on the viscosity of
PVC has also been studied [20]. The decrease in viscosity of the PVC matrix by addition
of low molecular weight polymers depends on their compatibility and binding force with
PVC. A different approach has been used to understand the melt behavior of PVC by
considering PVC as a fluid which contains filler (PVC primary particles) [22]. It has
been shown that fully fused PVC has lower viscosity than PVC powder (partially fused).
This phenomenon was explained as analogous to the behavior of fluid which contains
filler, and reduction in viscosity at fusion was attributed to the reduction in filler level as
the PVC primary particles melt out at higher temperatures.

Hayes [24] studied the effect of two lubricants, a process aid, and an impact

modifier on rheological properties of PVC using a central composite design. All studied

13

factors had a significant effect on the rheological properties of PVC. PVC in its melted
state is a non-Newtonian fluid with shear thinning behavior, also known as psuedoplastic
in nature. As a result, an increase in shear rate will decrease the viscosity of the PVC
melt. It was found that the increase in impact modifier and process aid content resulted
in increased non-Newtonian behavior of the PVC melt. Opposite effects were observed
between the two different lubricants studied (wax and oxidized polyethylene). Wax
decreased the non-Newtonian behavior while oxidized polyethylene increased the
behavior. This was attributed to the effect of these lubricants on the fusion behavior of
PVC. Oxidized polyethylene promotes a higher degree of fusion compared to wax,
which results in more entanglements of polymer chains, consequently increasing the non-
Newtonian behavior of the melt.

The incorporation of filler into polymer matrices complicates rheological
measurements of the system. The effect of filler on different rheological properties of a
polymer melt is a complex phenomenon and depends on several factors such as content,
shape and size of filler, interaction between filler and polymer, etc. Several researchers
have investigated the rheological properties of natural fiber filled polymer [25-28]. Maiti
and Hassan [28] studied the rheological properties of PP/wood-flour composites in a
capillary rheometer. They found that increasing the wood flour content increases the
viscosity and shear thinning behavior of the PP melt, whereas it reduces the melt
elasticity of the matrix. It was also observed that the viscosity of filled polymer is less
sensitive to temperature changes than that of unfilled polymer. Similar results were

observed by Xiao and Tzoganakis [26] for an HDPE/wood-flour composite system

14

except for melt elasticity, where the opposite trend was observed (increased melt
elasticity with increased wood content).

Recently, Guffey and Sabbagh [29] used chlorinated polyethylene (CPE) as a
compatibilizer for PVC/wood-flour composites and showed that addition of CPE to
PVC/wood-flour composites reduced the viscosity of the composites, resulting in
increased processability of the composites and higher outputs. The melt viscosity of
plasticized PVC is significantly increased when wood flour is added [30]. Increased
viscosity hinders bubble formation and growth during foaming and generally results in
poor foam quality in terms of average cell size and cell-population density. In contrast,
the blowing gas reduces the viscosity of the molten polymer due to the plasticization
effect on the polymer. The viscosity of the polymer matrix is a critical parameter for
bubble formation and growth during the foaming process; therefore it needs to be studied
in depth in order to understand how the wood flour, plasticizer, foam modifier and gas
content will affect the rheology and thus the foamability of the PVC composites.

Foaming of PVC WPCS addresses some of the drawbacks of these materials by
reducing the density and improving the end product properties. Unfortunately, this does
not solve another major issue related to incompatibility between hydrophilic wood and

hydrophobic PVC.

15

Coupling Agents for PVC/Wood—Flour Composites:

To achieve maximum reinforcement benefits from a filler used in a polymer, the
filler, which is strong and stiff, should bear most of the load or stress applied to the
composite system. Conversely, the polymer, which is of low strength, fairly tough and
flexible, should effectively transmit the load to the filler. To effectively transfer the load
from matrix to filler, the matrix must have strong interfacial adhesion with the filler.
Weak interfacial bonding will result in fiber pullout from the polymer matrix on
application of load, and will reduce the overall strength of the composite. Thus, it is the
interface which controls the final strength properties of the composite.

Although WPCS are superior to neat polymers in terms of material cost and
stiffness, their strength performance (tensile, flexural, and impact) is generally lower than
the unfilled polymers. A general trend of increasing modulus and decreasing strength
properties is observed as more wood is added to the polymer. Increased modulus by
addition of wood is a result of the high stiffness of the filler compared to the polymer,
while decreased strength is likely a result of the natural incompatibility of phases during
the mixing of the hydrophilic wood fibers with the hydrophobic polymer matrix [31-33].
Phase incompatibility yields very weak interactions and thus a weak interface between the
fiber and the matrix.

One approach to designing WPCS is to modify the wood fiber surface with
coupling agents to improve the strength. Coupling agents convert the hydrophilic surface
of wood fibers to a more hydrophobic state. Thus, the surface tension of wood fibers is

reduced and approaches that of the molten polymer. As a result, wetting and adhesion are

16

improved via mechanisms such as diffusion and mechanical interlocking between treated
fibers and the polymer matrix [33].

Due to its strong effect in altering the hydrophilic surface of wood fibers to a more
hydrophobic one, maleic anhydride functionalized polyolefin is commonly used as a
coupling agent for polyolefin/wood-fiber composites [34, 35]. Similarly, to enhance the
interfacial adhesion between wood fibers and a PVC matrix, several investigators have
assessed the effects of various fiber treatments, including different types of isocyanates,
maleic anhydride, linoliec acid, silanes, etc. as coupling agents [29, 36-38].

Kokta et al. [38] used linoleic acid, abietic acid and maleic anhydride as coupling
agents for composites of PVC with chemi-thermomechanical pulp of aspen. All of the
coupling agents showed increased tensile strength of the composites, compared to
composites without coupling agent. However, the strength properties of coupled
composites were still much lower than unfilled polymer. Linoleic acid treated
composites showed the highest tensile strength of the three coupling agents studied.
Interestingly, when aspen sawdust was used instead of chemi-thermomechanical pulp as a
filler for PVC with maleic anhydride as a coupling agent, the treated composite showed
reduced strength properties compared to untreated composites.

Most mechanical properties of the composites were improved by these chemical
treatments compared to those of composites with non-treated fibers. However, the
properties of the composites were inferior to those of the unfilled PVC, suggesting that,
unlike polyolefin/wood-fiber composites, the well-known method of converting the
hydrophilic surface of wood fiber to a hydrophobic one is not effective for enhancing the

adhesion of PVC to wood fibers.

17

Matuana et al. [39] demonstrated that when PVC is used as a matrix in WPCS,
acid-base interactions, in which one phase acts as an electron donor (base) and the other
acts as an electron acceptor (acid), are a significant factor in enhancing interfacial
adhesion. Therefore, surface modification of wood fibers to be used with PVC should be
designed to modify the acid-base interactions at the matrix/fiber interface in order to
improve the performance of these composites. For example, by changing the acidic
characteristics of wood fibers through surface modification with 'y-amino propyltriethoxy
silane, PVC/wood-fiber composite with equal tensile strength and greater modulus than
unfilled PVC was developed. The aminosilane successfully modified the wood surface,
and facilitated the interaction between the wood and PVC according to Lewis acid-base
theory [39].

In spite of these benefits, y—amino propyltriethoxy silane has not been extensively
used as a coupling agent for PVC/wood-fiber composites, mainly due to its high cost but
also due to the difficulty in evenly coating the surface of wood fibers, owing to its
tendency to hydrolyze and self-condense. Consequently, aminosilane is not a desirable
coupling agent in this application. Thus, one of the objectives of this study is to explore
the use of amine-containing natural polymers such as chitin and chitosan as cost-effective
and novel coupling agents for PVC/wood-fiber composites. The acetylamine
functionality of chitin, and the amine functionality of the chitosan, should permit these
polymers to interact with wood and PVC in a manner similar to the aminosilane, and so

enhance the interfacial adhesion between PVC and wood fibers.

18

Chitin and chitosan are natural, biodegradable and biocompatible materials which
are desirable for a variety of applications. Chitin is the second most abundant
polysaccharide in nature, following cellulose. Chitosan is produced by deacetylating
chitin under alkaline conditions, usually with sodium hydroxide. The structural
difference between chitin and chitosan has many consequences. For example, chitin is
not water soluble, while chitosan is soluble at acidic pH. Chitosan is also much more
reactive than chitin, due to the primary amine in the structure.

Several investigators have reported the use of chitin and chitosan with different
polymers for various applications [40-46]. In one study, the interfacial adhesion between
chitin fibers and polycaprolactone (PCL) was increased by an irradiation treatment of the
composites. The treatment showed an overall increase in mechanical properties of the
composite compared to composite prepared from untreated chitin fiber. This increase in
interfacial bonding was attributed to a free-radical grafting reaction [40]. Chitosan has
also been reported to have been crosslinked to a polymer matrix [41]. That process used
formaldehyde, which is a known carcinogen and hazardous to the environment, as a cross
linking agent.

Souza Rosa and Andrade prepared biodegradable composites by incorporating
chitin flakes ranging from 0-30 wt% into a plasticized starch matrix using an injection
molding process. Chitin flakes increased the elastic modulus, tensile stress and water
resistance of the composites when compared to the unfilled starch [42]. Biodegradation
of synthetic polymers was shown to increase by incorporating a natural biodegradable

polymer such as chitosan [43]. Thwe and Liao showed that the treatment of glass and

19

bamboo fiber with chitosan solution can improve the adhesion and properties of the
polypropylene hybrid composites [44].

Chitosan is also used in the wood industry. For example, chitosan forms a Schiff
base when reacted with aldehyde compounds. This property of chitosan can be very
useful in reacting with the formaldehyde released from the glue line of plywood, thus
reducing the overall emission of formaldehyde to the environment. Hence, the chitosan
can be used as a functional coating reagent for wood [45]. The use of chitosan as an
environmentally fiiendly adhesive for wood has also been reported in the literature. Glue
made from chitosan showed excellent water resistance and was proposed as a
replacement for synthetic adhesives [46]. The use of chitin and chitosan as coupling
agents for wood composites has not been reported in the literature.

Development of an efficient, cost effective method of introducing foamed
structure into PVC via a continuous microcellular foaming process using supercritical
fluids will reduce the weight and material cost of the composite while increasing various
properties. Similarly, improvements in strength properties can be achieved with a
coupling agent. Both of these developments would broaden the application base for PVC
wood composites. However, certain segments of the composites market are beginning to
look for alternatives to synthetic polymers. As biopolyrners have emerged as a viable,
environmentally friendly alternative to synthetic polymers, natural polymers such as PLA

are finding use in several applications.

20

Polylactide:

Polylactide (PLA) is a renewable and biodegradable linear polyester formed from
lactic acid (see Appendix D for additional information). PLA offers high stiffness, easy
processability and excellent clarity and gloss. This makes PLA 3 better alternative to
replace synthetic commodity polymers in many applications. Products such as packaging
and some automotive parts have relatively short life spans compared to materials used in
the building and construction industry and would be ideal markets for PLA.

Natural fiber provides a low cost, biodegradable, stiff and lightweight
reinforcement for PLA, which has a disadvantage of higher density than most commodity
polymers such as PE, PP and PS. The melting point of PLA is around 165°C, which also
makes it suitable to use with natural cellulosic fibers. Oksman and coworkers [47] have
shown that PLA/flax fiber composites have superior mechanical properties compared to
PP/flax fiber composites, and therefore have the potential to replace PP/natural fiber
composites in many industrial applications. PLA composites have also been
demonstrated with cellulose [48], newsprint [49] and bamboo fibers [50]. These results
imply that wood, which is comparable in composition to other natural fibers, could also
be used in making PLA composites. Wood flour is abundantly available in the US as a
waste product of other manufacturing processes such as lumber production and fumiture
manufacture, which makes the use of wood flour economical and easy to implement.
Thus, another goal of this study is to evaluate the use of wood flour as a filler for PLA.

Oksman et al. have reported the need for better interfacial adhesion between the
PLA matrix and natural fibers to improve overall mechanical properties of the

composites. Huda et al. [51] found that addition of 30 wt. % cellulose fibers reduces the

21

flexural strength of the composite while increasing the modulus by two folds over
unfilled PLA polymer. Similar results were obtained by other researchers when bamboo
[50] and recycled newsprint fibers [49] were used. These results show there is weak
interfacial bonding between the reinforcing fibers and polymer matrix. The strength of
natural fiber composites depends on the fiber dispersion and the interfacial bonding
between the fibers and the polymer matrix. Poor dispersion is inherent to cellulosic
materials due to hydrogen bonding between the fibers [31, 52, 53]. Different approaches
have been used to overcome this problem. In one case, bamboo fiber itself was esterified
with maleic anhydride and successfully used as a compatibilizer for PLA/bamboo fiber
composites [50]. Addition of 5% esterified bamboo fiber resulted in around 21%
improvement in the tensile strength and 10% increase in tensile modulus, while further
addition of compatibilizer had no significant effect on the properties. This increase in the
properties is an indication of increased interaction between the PLA matrix and bamboo
fiber. Even though bamboo fiber esterified with maleic anhydride increased the
properties of the PLA composite, this approach has several drawbacks, including the
additional step of manufacturing reactive fibers before processing of the composites, and
the use of a solvent process to produce reactive fibers, which is less desirable in the
industry as well as uneconomical. Another approach is to use maleated PLA as a
coupling agent. Initial results suggest that this would be an effective coupling agent for
PLA/natural fiber composites [54]. However, this coupling agent is not available
commercially and would be expensive and time consuming to produce as a part of a
composite manufacturing process. A commercially available natural product is desirable

as a coupling agent for PLA cellulosic fiber composites. Chitosan is natural,

22

biodegradable and readily available as a consumer product, and has been shown to have
compatibility with wood. An additional goal of this work is to investigate the use of
chitosan as a coupling agent for PLA wood flour composites that could improve the

interfacial adhesion between hydrophilic wood and the hydrophobic PLA matrix.

23

Summary

WPCS have gained tremendous attention in recent years, and the market continues
to grow. However, several drawbacks need to be addressed before they can reach their
maximum potential. Microcellular foaming of PVC wood composites addresses some of
the drawbacks by not only reducing the weight and material cost of the composite but
also increasing numerous composite properties. Although microcellular foaming of PVC
and its composites has been successfully demonstrated by the batch process, a continuous
process is needed to make it commercially viable. Continuous extrusion using
supercritical fluids would be a cost effective, efficient way to produce these foams. In
order to control the morphology and tailor the end product properties of microcellular
foamed PVC composites, various factors that control the rheology of the polymer must be
investigated. Specifically, the complex interactions between wood flour and other
additives must be understood.

Incompatibility between hydrophilic wood and hydrophobic polymer matrix leads
to poor strength properties of the composites. Coupling agents are routinely used to
overcome this issue. However, there has been only limited success with coupling agents
for PVC wood composites. An effective, low cost coupling agent is still needed for PVC
wood composites.

Biopolymers such as PLA represent a growing alternative to synthetic polymers.
PLA natural fiber composites could be developed for use in many applications.
Biopolymers like PLA which offer comparable properties to synthetic polymers while
being renewable, sustainable and biodegradable would be preferred for use in products

with relatively short life spans. Results of other studies suggest that wood can be used as

24

a filler for PLA composites, but this must be investigated to determine whether PLA
WPCS would have sufficient properties for use in various applications. Regardless of the
cellulosic based filler used, a natural and cost effective coupling agent is needed to

improve the interfacial adhesion between the filler and PLA.

25

HYPOTHESES

This research is focused on improving WPCS properties and broadening their use

for a variety of applications. End results of this project are expected to advance wood

composite technology by providing both fundamental and applied knowledge that can be

used to improve WPCS performance. The following hypotheses comprise the main focus

of this project.

1.

Material composition (wood flour, plasticizer and foam modifier contents) will
have a significant effect on the rheological properties of the PVC matrix, which
will affect both the processability of the PVC/wood-flour composites and the cell

nucleation and growth process during foaming.

Extrusion processing conditions (gas content, die temperature and extrusion
speed) and material composition variables (wood flour and foam modifier
content) will significantly affect the cell morphology of foamed PVC samples,
thus these factors must be investigated to understand the dynamics of bubble

nucleation, growth and stabilization during a continuous extrusion process.

Chitin and chitosan can effectively increase the interfacial adhesion between PVC
and wood flour and thus can be used as effective coupling agents for PVC/wood-

flour composites.

Wood flour will be a suitable filler for PLA in natural, biodegradable composites.
Further, chitosan can be used as a natural coupling agent to overcome the natural
incompatibility between hydrophilic natural fibers and the hydrophobic PLA

matrix.

26

10.

REFERENCES

Matuana, L.M., Park, CB, and Balatinecz, J.J., "Cell morphology and property
relationships of microcellular foamed PVC/wood-fiber composites", Polymer
Engineering and Science, 38(11): 1862-1872 (1998).

Matuana, L.M., Park, CB, and Balatinecz, J .J ., "Processing and cell morphology
relationships for microcellular foamed PVC/wood-fiber composites", Polymer
Engineering and Science, 37(7): 1137-1147 (1997).

Matuana-Malanda, L., Park, CB, and Balatinecz, J.J., "Characterization of

microcellular foamed PVC/cellulosic-fiber composites", Journal of Cellular
Plastics, 32(5): 449-469 (1996).

Mengeloglu, F. and Matuana, L.M., "Foaming of rigid PVC/wood flour
composites through a continuous extrusion process", Journal of Vinyl & Additive
Technology, 7(3): 142-148 (2001).

Mengeloglu, F ., Matuana, L.M., and King, J .A., "Effects of impact modifiers on
the properties of rigid PVC/wood-fiber composites", Journal of Vinyl & Additive
Technologi, 6(3): 153-157 (2000).

Matuana, L.M. and Mengeloglu, F ., "Microcellular foaming of impact-modified
rigid PVC/wood-flour composites", Journal of Vinyl & Additive Technology, 7(2):
67-75 (2001).

Doroudiani, S., Park, CB, and Kortschot, M.T., "Characterization of
microcellular foamed HDPE/PP blends", Annual Technical Conference - Society
of Plastics Engineers, 54(Vol. 2): 1914-1919 (1996).

Rachtanapun, P., Selke, S.E.M., and Matuana, L.M., "Microcellular foam of
polymer blends of HDPE/PP and their composites with wood fiber", Journal of
Applied Polymer Science, 88(12): 2842-2850 (2003).

Kumar, V. and Weller, J.E., "A process to produce microcellular PVC",
International Polymer Processing, 8(1): 73-80 (1993).

Kweeder, J.A., Ramesh, N.S., Campbell, G.A., and Rasmussen, D.H., "The
nucleation of microcellular polystyrene foam", Annual Technical Conference -
Society of Plastics Engineers, 49: 1398-400 (1991).

27

ll.

12.

13.

14.

15.

16.

l7.

l8.

19.

20.

21.

Kumar, V. and Weller, J.E., "Microcellular polycarbonate - Part 1: Experiments
on bubble nucleation and growth", Annual Technical Conference - Society of
Plastics Engineers, 49: 1401-5 (1991).

Matuana, L.M., Park, C.B., and Balatinecz, J.J., "Structures and mechanical
properties of microcellular foamed polyvinyl chloride", Cellular Polymers, 17(1):
l-16(l998).

Baldwin, D.F., Park, C.B., and Sub, N.P., "An extrusion system for the processing
of microcellular polymer sheets: shaping and cell growth control", Polymer
Engineering and Science, 36(10): 1425-1435 (1996).

Baldwin, D.F., Park, C.B., and Sub, N.P., "A microcellular processing study of
poly(ethylene terephthalate) in the amorphous and semicrystalline states. Part 11.
Cell growth and process design", Polymer Engineering and Science, 36(11):
1446-1453 (1996).

Behravesh, A.H., Park, C.B., and Venter, R.D., "Extrusion of low-density
microcellular HIPS foams using C02", M) (American Society of Mechanical
Engineers), 76(Cellular and Microcellular Materials): 47-63 (1996).

Park, C.B., Pan, M., and Yeung, A.K. "Microcellular processing of low-density
PS foams in extrusion". in Cellular Polymers IV, International Conference, 4th,
Shrewsbury, UK, June 5-6, 1997.

Matuana, L.M. and Mengeloglu, F., "Manufacture of rigid PVC/wood-flour

composite foams using moisture contained in wood as foaming agent", Journal of
Vinyl & Additive Technology, 8(4): 264-270 (2002).

Dey, S.K., Jacob, C., and Xanthos, M., "Inert-gas extrusion of rigid PVC foam",
Journal of Vinyl & Additive Technology, 2(1): 48-52 (1996).

Collins, EA. and Krier, C.A., "Poly(vinyl chloride) melt rheology and flow
activation energy", Transactions of the Society of Rheology, 11(2): 225-42 (1967).

Sieglaff, C.L., "Rheological behavior of poly(vinyl chloride) mixtures. l. Viscous
behavior", Polymer Engineering and Science, 9(2): 81-5 (1969).

Glomsaker, T., Hinrichsen, EL, and Thorsteinsen, P., "In-line versus off-line
rheological measurements on S-PVC formulations releant for flow in extrusion
dies", Polymer Engineering & Science, 41(12): 2231-2248 (2001).

28

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

Berard, M.T., "Fusion mechanism in rigid PVC", Journal of Vinyl & Additive
Technology, 8(4): 246-250 (2002).

Shah, P.L., "Significance of capillary rheology testing temperture for predicting
PVC extrudability", Polymer Engineering and Science, 14(11): 773-7 (1974).

Hayes, V.O., "Effects of formulation variables on rheology of rigid PVC",
Journal of Vinyl & Additive Technology, 5(2): 71-75 (1999).

George, J., Janardhan, R., Anand, J.S., Bhagawan, SS, and Thomas, S., "Melt
rheological behaviour of short pineapple fibre reinforced low density
polyethylene composites", Polymer, 37(24): 5421-5431 (1996).

Xiao, K. and Tzoganakis, C., "Rheological properties and their influence on
extrusion characteristics of HDPE-wood composite resins", Annual Technical
Conference - Society of Plastics Engineers, 60th(Vol. 1): 252-256 (2002).

Marcovich, N.E., Reboredo, M.M., Kenny, .l., and Aranguren, M.I., Rheology
Acta, 43: 293 (2004).

Maiti, SN. and Hassan, M.R., "Melt rheological properties of polypropylene-
wood flour composites", Journal of Applied Polymer Science, 37(7): 2019-2032
(1989)

Guffey, V0. and Sabbagh, A.B., "PVC/wood-flour composites compatibilized
with chlorinated polyethylene", Journal of Vinyl & Additive Technology, 8(4):
259-263 (2002).

Goettler, LA, "The extrusion and performance of plasticized poly(vinyl chloride)
hose reinforced with short cellulose fibers", Polymer Composites, 4(4): 249-255
(1983)

Saheb, D.N. and Jog, J.P., "Natural fiber polymer composites: A review",
Advances in Polymer Technology, 18(4): 351-363 ( 1999).

Bledzki, AK. and Gassan, J., "Composites reinforced with cellulose based
fibres", Progress in Polymer Science, 24(2): 221-274 (1999).

Stokke, DD. and Gardner, D.J., "Fundamental aspects of wood as a component of
thermoplastic composites", Journal of Vinyl and Additive Technology, 9(2): 96-
104 (2003).

29

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

Hristov, V. and Vasileva, S., "Dynamic mechanical and thermal properties of
modified poly(propylene) wood fiber composites", Macromolecular Materials
and Engineering, 288(10): 798-806 (2003).

Woodhams, R.T., Thomas, G., and Rodgers, D.K., "Wood fibers as reinforcing
fillers for polyolefins", Polymer Engineering & Science, 24(15): 1166—1171
(1984)

Kokta, B.V., Maldas, D., Daneault, C., and Beland, P., "Composites of poly(vinyl
chloride) - wood fibers. 111: Effect of silane as coupling agent", Journal of Vinyl
Technology, 12(3): 146-53 (1990).

Kokta, B.V., Maldas, D., Daneault, C., and Beland, P., "Composites of poly(vinyl
chloride)-wood fibers. 1. Effect of isocyanate as a bonding agent", Polymer-
Plastics Technology and Engineering, 29(1-2): 87-118 (1990).

Kokta, B.V., Maldas, D., Daneault, C., and Beland, P., "Composites of poly(vinyl
chloride) and wood fibers. Part II: Effect of chemical treatment", Polymer
Composites, 11(2): 84-9 (1990).

Matuana, L.M., Balatinecz, J .J ., and Park, C.B., "Effect of surface properties on
the adhesion between PVC and wood veneer laminates", Polymer Engineering

and Science, 38(5): 765-773 (1998).

Yang, A. and Wu, R., "Enhancement of the mechanical properties and interfacial
interaction of a novel chitin-fiber-reinforced poly(e-caprolactone) composite by
irradiation treatment", Journal of Applied Polymer Science, 84(3): 486-492
(2002)

Zhang, Q., Liu, L., Ren, L., and Wang, F., "Preparation and characterization of
collagen-chitosan composites", Journal of Applied Polymer Science, 64(11):
2127-2130 (1997).

Souza Rosa, R.C.R. and Andrade, C.T., "Effect of chitin addition on injection-
molded thermoplastic corn starch", Journal of Applied Polymer Science, 92(4):
2706-2713 (2004).

Ratajska, M. and Boryniec, S., "Biodegradation of some natural polymers in
blends with polyolefines", Polymers for Advanced Technologies, 10(10): 625-633
(1999)

30

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

Thwe, M.M. and Liao, K., "Tensile behaviour of modified bamboo—glass fibre
reinforced hybrid composites", Plastics, Rubber and Composites, 31(10): 422-431
(2002)

Sato, K., Ota, H., and Omura, Y., "Development of functional coating reagent for
wood based materials by using chitosan", Advances in Chitin Science, 2: 897-901
(1997)

Umemura, K., Inoue, A., and Kawai, 8., "Development of new natural polymer-
based wood adhesives I: dry bond strength and water resistance of konjac
glucomannan, chitosan, and their composites", Journal of Wood Science, 49(3):
221-226 (2003).

Oksman, K., Skrifvars, M., and Selin, J.F., "Natural fibres as reinforcement in
polylactic acid (PLA) composites", Composites Science and Technology, 63(9):
1317-1324 (2003).

Huda, M.S., Mohanty, A.K., Drzal, L.T., Schut, E., and Misra, M., ""Green"
composites from recycled cellulose and poly(lactic acid): Physico-mechanical and
morphological properties evaluation", Journal of Materials Science, 40(16):

4221-4229 (2005).

Huda, M.S., Drzal, L.T., Misra, M., Mohanty, A.K., Williams, K., and Mielewski,
D.F., "A Study on Biocomposites from Recycled Newspaper Fiber and Poly(lactic
acid)", Industrial & Engineering Chemistry Research, 44(15): 5593-5601 (2005).

Lee, S.-H., Ohkita, T., and Kitagawa, K., "Eco-composite from poly(lactic acid)
and bamboo fiber", Holzforschung, 58(5): 529-536 (2004).

Huda, M.S., Mohanty, A.K., Drzal, L.T., Schut, E., and Misra, M., "\"Green\"
composites from recycled cellulose and poly(lactic acid): Physico-mechanical and

morphological properties evaluation", Journal of Materials Science, 40(16):
4221-4229 (2005).

Kazayawoko, M., Balatinecz, J .J ., and Matuana, L.M., "Surface modification and
adhesion mechanisms in wood fiber-polypropylene composites", Journal of
Materials Science, 34(24): 6189-6199 (1999).

Felix, J .M. and Gatenholm, P., "The nature of adhesion in composites of modified
cellulose fibers and polypropylene", Journal of Applied Polymer Science, 42(3):
609-620 (1991).

31

54. Plackett, D., "Maleated Polylactide as an Interfacial Compatibilizer in
Biocomposites", Journal of Polymers and the Environment, 12(3): 131-138
(2004).

32

CHAPTER 3

ON LINE MEASUREMENT OF RHEOLOGICAL PROPERTIES OF
PVC/WOOD-FLOUR COMPOSITES*

This chapter was expanded from the paper published in the Journal of Vinyl and Additive
Technology, 10(3), 121-126, 2004.

‘This paper was the winner of the 2004 Best Student Paper Award presented by the SPE
Vinyl Division

33

ABSTRACT

Using a factorial design approach, this study examined the effect of the
component materials on the viscoelastic properties of PVC/wood-flour composites.
Statistical analysis was performed to determine the effects of wood flour content, acrylic
modifier content and plasticizer content on the die swell ratio and viscosity of the
composites measured online on a conical twin-screw extrusion capillary rheometer. The
visco-elastic properties of the samples were also measured using dynamic mechanical
analysis (DMA). Wood flour content and acrylic modifier content were the two
important variables affecting the die swell ratio, whereas the addition of a low level of
plasticizer did not affect the die swell ratio. The die swell increased with the increased
acrylic modifier content but it was considerably reduced by adding wood flour into the
PVC matrix. The true viscosity of neat PVC and PVC/wood-flour composites decreased
with the plasticizer content, irrespective of the acrylic modifier content. However, the
addition of acrylic modifier significantly increased the viscosity of unfilled PVC while an
opposite trend was observed for the composites due to the differing effect of acrylic

modifier on the melt elasticity and viscosity of these materials.

Keywords: PVC/wood-flour composites, die swell ratio, viscosity, extrusion capillary

rheometer, plasticizer, acrylic modifier.

34

INTRODUCTION

The United States of America has the largest market for both unfoamed and
foamed wood-plastic composites (WPCS) in the world, with an expected annual grth
rate of 12 % [1]. The enormous advantages of unfoamed and foamed WPCS over solid
wood like the low maintenance, high strength to weight ratio, no splintering, chemical
resistance, water resistance and good weatherability make them more suitable for
stringent applications ranging from decking to automotive parts [1-5].

Polyvinyl chloride (PVC), being the second largest plastic used in manufacturing
WPCS, preceded by polyethylene, has limitations due to its complex processability,
which worsens in combination with wood flour. The high melt viscosity of PVC/wood-
flour composites, due to the incorporation of wood flour into the PVC matrix, may lead
to poor quality products or restrain the bubble formation and growth during the foaming
process. This can prevent the formation of low density foamed composite products with
uniform and homogeneous cell structures [6].

Selecting a proper processing aid to get a better quality product with enhanced
processability is a key factor in manufacturing [8]. With appropriate addition level, the
incorporation of acrylic processing aid is known to promote fusion of the PVC matrix
with an increase in melt strength and melt elasticity. However, excessive addition of
acrylic processing aid may change the rheology of the resin resulting in melt fracture [7].
Acrylic processing aid (foam modifier) has been used for several years to modify the
PVC matrix to ease the foaming process. Matuana and Mengeloglu [2] have reported

that the addition of acrylic modifiers, even at a low level, can significantly affect the melt

35

viscosity of the filled PVC matrix, thus enhancing the foamability of PVC/wood-flour
composites. Processing of both unfoamed and foamed PVC/wood-flour composites can
also be enhanced by the addition of plasticizer into the formulation due to the softening
effect of plasticizer [9].

Several researchers have characterized the rheological properties of neat PVC
[10-14]. Collins and Krier [10] have found two different activation energies for PVC. A
low activation energy for a low temperature region characterized by particulate flow and
a high activation energy for the high temperature region characterized by molecular flow
was observed. The temperature at which the transition from the high to the low
temperature region occurred was a function of shear stress. The effects of low molecular
weight polymers like acrylate and polypropylene terephthalate on the viscosity of PVC
have also been studied [1 1]. The decrease in viscosity of the PVC matrix by addition of
low molecular weight polymers depends on their compatibility and binding force with
PVC. A different approach has even been used to understand the melt behavior of PVC
by considering PVC as a fluid which contains filler (PVC primary particles) [13]. It has
been shown that fully fused PVC has lower viscosity compared to PVC powder (partially
fused). This phenomenon was explained as analogous to the behavior of fluid which
contains filler, and reduction in viscosity at fusion was attributed to the reduction in filler
level as the PVC primary particles melt out at higher temperatures.

It is of immense importance to know the rheological behavior of polymer at
processing conditions during extrusion to predict or tailor the extrudability of the polymer
and the surface quality of the extrudate. Compared to a traditional capillary rheometer,

an extrusion capillary rheometer generates the test results at the processing conditions,

36

which gives it a wide applicability. Glomsaker et al. [12] have analyzed and compared
the in-line (extrusion capillary rheometer) versus off-line (capillary rheometer)
rheological data of PVC for better simulation of flow in extrusion dies. Different
formulations showed different comparisons of in-line data with off-line. Rheological
data measured in the capillary rheometer are useful in predicting the thermal stability and
the processability of PVC [15,16].

The rheological properties of natural fiber filled polyolefins have been
investigated and reported by a number of researchers [l7-20]. Incorporation of natural
fibers in polyolefins increases the melt viscosity of the matrix and decreases the die swell
ratio of the compound [17,20]. However, much less attention has been paid to evaluating
the rheological properties of PVC/wood-flour composites; in particular the effect of
various additives on the rheology of PVC/wood-flour composites has not been studied
thoroughly. The addition of cellulose fibers into the plasticized PVC matrix increases the
viscosity of the composite compared to neat PVC [21]. Recently, Guffey and Sabbagh
[22] used chlorinated polyethylene (CPE) as a compatibilizer for PVC/wood-flour
composites and showed that addition of CPE to PVC/wood-flour composites reduced the
viscosity, resulting in increased processability of the composites and higher outputs.

The rheological characteristics of the melt are the most vital aspects in
determining not only the surface quality of unfoamed products but also the bubble
formation and growth during the foaming process, and need to be studied in detail. The
main objective of this study was to investigate the effect of different formulation
variables on the rheological properties (die swell and true viscosity) of PVC/wood-flour

composites using a factorial design approach. A two-level factorial design was selected

37

in this study because it is a valuable approach for examining the effect of several
variables in the system at one time, in addition to revealing the effects of interactions
between variables that cannot be obtained from the traditional method of changing one

variable at a time [23].

38

EXPERIMENTAL

Materials

PVC (K value-66) resin was supplied by Oxyvinyls. Tin stabilizer (PlastiStab
2808) used was from OMG Americas. Calcium stearate (Synpro, Ferro Corp.) and
paraffin wax (Gulf Wax) were used as lubricants. Di-2-ethylhexyl phthalate (DEHP)
plasticizer (Aristech Chemical Corp.) and acrylic processing aid (foam modifier)
(Paraloid K-400, Rohm and Haas Co.) were also used in the formulations. Wood flour
from hardwood maple species (40-mesh size — 425 um) supplied by American Wood
Fibers (Schofield, Wisconsin) was used as a filler. Table 3.1 summarizes the typical

formulations used in the experimental design.

39

Table 3.1. Formulations Used in Neat PVC and PVC/Wood-Flour Composites.

 

 

Ingredients Concentration (phr)
PVC K-66 100
Tin stabilizer 2
Calcium stearate 1.5
Paraffin wax 2
Acrylic modifier (K-400) 0 — 8
Wood flour 0 — 50
Di-2-ethylhexyl phthalate (DEHP) 0 - 4

 

40

Experimental Design and Compounding

A two-level factorial design with center point was developed to study the
rheological properties of PVC/wood-flour composites using Design Expert® software
v.6.0 (Stat-Ease Corp., Minnesota). Wood flour content, acrylic modifier content and
plasticizer content were the three key variables studied, whereas the die swell ratio and
true viscosity were the variable responses obtained in this study. The generated
experimental design is summarized in Table 3.2.

The Design Expert® software provides two different methods of displaying the
factor levels in the experimental design: (i) the actual levels of the factors (i.e., the actual
values in the experiment) and (ii) the coded factor levels (i.e., -l for low levels, +1 for

high levels and 0 for center point). The coded factor levels are defined as [6,19]:

 

Coded factorlevels= actual value- factor mean (1)
(range of the factonalvalue/Z)

Coding reduces the range of each factor to a common scale, -I to +1, regardless of
its relative magnitude. It is also easier to think in terms of changes from low to high for
the factors than to think about their actual values. The experimental design matrix in
terms of both actual and coded factor levels is shown in Table 3.2. Six replicates were
used for each experiment to determine the statistical significance of the results.
Relationships between the wood flour content, acrylic modifier content and plasticizer
content and the responses (die swell ratio and true viscosity) were statistically modeled to

better understand the significant factors and interactions in the system.

41

$39» covoo 3a momofinummm E mega?“

 

 

 

42

0.5 32 3mm mam 8v m 8C 2. 8V 2 E80 a
3mm 32 4.2“ m3 st 2. a: m :t om Em w
3m 0.8m :2 m: st 4 at w m3 0 Em m
£2 3% m E mom 2: 4 :2 0 ct om Em o
0; Pa 2: Rm at 4 A: o i o Em m
22 an: 3.4“ m3 A: o at m are om Em 4
30m 28. 2mm :2 A: o at w A: o Em m
32 mmmr 58.... Sm i o :2 o a: 0m. Em m
mm: mo: 32 w: A: o A: o :2 o Em m
:e 82 re 82 9e 8m 3:: 33 33 33
Amémv . =95. omn— 3355 38:3 2.2.3... awn. moan—=2
moan.— ..aeam 398.5% .a 3763.5 oF—H hiatus—m 521.338 92.934 .50.» .50.?» magi—Hm

 

mom—3&3“ unhouonm

 

.EEmom atomumaaan .3 3.228 “.95.. 3.3m 3.30 ea .53. mo wash a use». sumac 35.3.2.5 .3 «Sam

Before compounding, wood flour was dried in an oven for 48 hrs at 105°C to
remove the moisture. The ingredients were weighed at the addition levels described in
the experimental design (Table 3.2) and mixed in a high intensity batch mixer
(Papenmeier, TGAHK20) for 10 min. It should be mentioned that a low level of
plasticizer was used throughout the experiments to remain in the antiplasticization region
where both neat PVC and PVC/wood-flour composite samples exhibit greater stiffness
and strength compared to unplasticized counterparts [9]. The compounded batch was

then used for rheological tests.

Rheology Experiments

The rheology of the compounded batch was measured online on an Intelli-Torque
Plasticorder® Torque Rheometer (C. W. Brabender Instruments Inc.) equipped with a 32
mm conical counter-rotating twin-screw extruder (L/D ratio 13:1). Different L/D ratios
(10:1 and 15:1) of capillary inserts were used. The extruder’s barrel temperature profile
and die temperature were kept constant at 190°C to maintain a constant melt temperature
for accurate viscosity measurements. However, the speed of the screws was varied from
5 rpm to 30 rpm to generate different shear rates. The collected extrudates were used to
measure the die swell ratio, which was calculated as the ratio of the solid extrudate

diameter to the diameter of the die.

The apparent shear stress (1,) in a capillary die is given by [24,25]:

1 _ AP
a MUD)

 

(2)

where AP is the pressure drop across the capillary die and L/D is the length to diameter

ratio of the die.

43

The apparent shear rate (ya ) was calculated using the following equation [24,25]:

- 4° (3)

Ya - nR3

 

where Q is the volumetric flow rate of the polymer melt and R is the radius of the
capillary.

The Bagley correction was applied to the apparent shear stress to account for the
excess pressure drop at the capillary entry. This correction converts apparent shear stress
to true wall shear stress. Measurements of viscosity at the same shear rate, with at least
two different capillaries are needed to apply the Bagley correction. Data of total pressure
drop versus flow rate are collected for each capillary die. The pressure drop versus L/D
data are plotted at each flow rate and lines are extrapolated to L/D equals to zero. The
resulting pressure drop (APC) is the pressure loss due to entrance effect at the particular

flow rate and die diameter. The Bagley corrected shear stress was calculated by [24, 25]:

1' : ALA—1:6. (4)
w 4(L/D)

where APc is the pressure drop at the capillary entrance.
As the apparent shear rate gives only Newtonian behavior (constant viscosity), a
correction is necessary for shear thinning fluids (pseudoplastic fluids). The Rabinowitsch

correction was done to calculate shear rate at the wall ( y w ) of the capillary die by using

the following equation [24, 25]:

 

. 1 .
rw=[3:: J76, (5)

44

where n is the flow behavior index obtained as the slope of the linear plot of log ya vs.
log 1“,.

The true melt viscosity was then calculated from the corrected shear stress and
corrected shear rate by using the following equation:

"zr—W' (6)

W

This calculated true viscosity was used for further analysis.

Dynamic mechanical analysis was carried out on a Perkin Elmer (DMA 7e)
instrument in the three-point bending mode to determine the storage modulus (elastic
modulus) and loss modulus (viscous modulus) of neat PVC and composites. The test was
performed under a helium atmosphere at a frequency of 1 Hz and a heating rate of
5°C/min in a temperature range from 25 to 150°C. The samples were cut from a panel
made by compression molding the extrudates at 190°C for 5 min. Sample dimensions

were 15 mm x 3 mm x 1.5mm thick. The average of three replicates was reported.

45

RESULTS AND DISCUSSION

Statistical Analysis of the Die Swell Ratio

Die swell is a phenomenon that occurs due to the elastic recovery of polymer after
shear deformation. This can be attributed to the memory effect of polymer chains, which
try to regain their original shape and size on stress removal.

A two-level factorial design was employed to evaluate the statistical effects of
wood flour content, acrylic modifier content and plasticizer content on the die swell ratio
of neat PVC and PVC/wood-flour composites. The analysis of the data was done to
identify the key factors affecting the die swell ratio and also to understand interactions
between different variables.

The die swell ratio data were entered into the design matrix (Table 3.2) generated
by Design-Expert® sofiware and a regression analysis was carried out to obtain the best-
fit model to the experimental data. A standard analysis of variance (ANOVA) showed
that the die swell data were best fitted with a linear model (values of "Prob > F" less than
0.05). When the values of "Prob > F" less are than 0.05, this indicates that the terms in
the model have a significant effect on the response. In other words, the probability of the
factors having no effect on the response is less than 5%. The ANOVA also showed that
curvature in the model was insignificant, i.e. the relationship was linear (see Appendix
B). The test for curvature compares the experimental data at the center point conditions
to the predicted values for that same point. Significant curvature can indicate quadratic
effects are present in the system, which are not accurately represented by the linear

equation. As the model contained both significant and non-significant terms, it was

46

simplified by dropping the insignificant term (plasticizer, term C). This modification did
not affect the adequacy of the model, which can be seen in Table 3.3 by reasonably good
R2 and adjusted R2 values as well as the adequate precision value of 19.6. R2 represents
the variation around the mean, adjusted R2 shows the variation around the mean adjusted
by the number of terms, and adequate precision is the signal to noise ratio for the model.
As a result, the derived linear regression equation describing the relationship between the

die swell ratio and different variables (in terms of coded factors) was reduced to:

Die swell ratio = +1.34 - 0.15A + 0.208 - 0.073AB (8)
where A is the coded variable that represents wood flour content and B is the coded

variable that represents acrylic modifier content.

Equation 8 indicates that wood flour content (factor A) and acrylic modifier
content (factor B) are the main effect factors, while wood flour content/acrylic modifier
content (interaction AB) is the only significant two—factor interaction term that affects the
die swell ratio. The effect of plasticizer (factor C) on die swell ratio was not significant
at the low addition level used in this study and was eliminated from the model.
Comparison of the coefficient and algebraic sign of each factor in Eq. 8 gives an
estimation of its relative influence on the die swell ratio of the sample.

As expected, wood flour content (factor A) had a negative effect on the die swell,
i.e., the die swell decreased as the concentration of wood flour increased in the
composites because of two main reasons: (i) reduction in normal stresses owing to low

shear strain in the capillary [24] and (ii) the addition of wood flour to plastics greatly

47

Table 3.3. Probability and Summary Statistics for Two-Level Factorial Model

 

 

Responses F- Value Prob > Fa R2 Adjusted R2 3:213:11:
Die swell ratio 79.2 0.0005 0.9834 0.9710 19.6
True “5°95.“ 651.7 0.0015 0.9994 0.9979 79.7

at5005
True V‘scoil‘y 1675.2 0.0016 0.9998 0.9992 122.6

at 1200 5
True viscosity

at 2000 3-. 3473.2 0.0003 0.9999 0.9982 166.8

 

alValues of “Prob>F” less than 0.05 indicates the model terms are significant.

bAdequate precision measures the signal to noise ratio. A ratio greater than 4 is
desirable for the model.

48

reduces the coefficient of thermal expansion of the matrix, which reduces the amount that
the polymer expands with applied heat [26]. By contrast, the die swell increased with the
acrylic modifier content (factor B) due to the increased melt elasticity [7].

The perturbation plot of die swell ratio against all three investigated variables
(Figure 3.1) provides supporting evidence of the relative importance of wood flour
content (factor A) and acrylic modifier content (factor B) on the die swell ratio of the
sample. Figure 1 illustrates changes in the die swell ratio as each factor moves from the
chosen reference, with all other factors held constant at the middle of the design space
(the coded zero level of each factor). It can be clearly seen from the perturbation graph
that wood flour content and acrylic modifier content were the two important variables
affecting the die swell, whereas the addition of a low level of plasticizer had no effect on
die swell. Since the two main factors affecting the die swell of the materials (factors A
and B) are also part of the only significant interaction (interaction AB) (Equation 8), it
would be inappropriate to investigate these factors individually because the effect of one
factor will depend upon the level of the other [23].

Figure 3.2 shows the interaction graph of the variation of die swell ratio as a
function of the interaction between wood flour content and acrylic modifier content. The
die swell ratio increased with the increased acrylic modifier content, regardless of the
addition level of wood flour. However, the extent of die swell ratio increase due to the
addition of acrylic modifier was more prominent for the unfilled PVC matrix than for the
composites. This behavior is clearly seen by comparing the magnitude of the difference
between the lowest (0 phr) and highest (8 phr) acrylic modifier content lines (Figure 3.2).

Without wood flour the difference is large (coded value -1 or neat PVC), while the

49

Perturbation

 

 

 

 

 

1.8-
1.6—
0 B
'23 A
i
, 1.41
U)
Q)
'5 C C
1.3—
1.1—
T l I l l
-10 05 0.0 05 10

Deviation from Reference Point

Figure 3.1. Perturbation plot of die swell against three investigated variables. A
represents the wood flour content; B is the acrylic modifier content;
and C is the plasticizer content.

50

 

1.8-

   

8me-400

l

1.6—

  

1.4—

Die swell ratio

OphrK-400

1.2- \j/

1.0-

 

 

 

] l l l I
-1.0 -0.5 0.0 0.5 1.0

Wood flour content

Figure 3.2. Die swell ratio as a function of the interaction between wood flour
content and at 0 and 8 phr acrylic modifier (K-400) content.

51

difference decreases with increased wood flour content (coded value +1 or composite
with 50 phr wood flour). The die swell ratio was considerably reduced by adding wood
flour into the PVC matrix. These results imply that the die swell ratio can be reduced by

increasing wood flour content in the PVC matrix.

Statistical Analysis of the True Viscosity

The same two-level factorial design was used to evaluate the effects of wood flour
content, acrylic modifier content, and plasticizer content on the true melt viscosity of the
neat PVC and PVC/wood-flour composites.

Since the shear rate encountered in the extruder during processing is in the range
of 102 to 103 s'l [25], melt viscosity was measured for each composition in the design
matrix at three different shear rates of 500, 1200 and 2000 s'l (Table 3.2). A normal
probability plot of the residuals for each of the shear rates indicated there was non-normal
variance. ANOVA requires equal variance of the residuals to detect differences in the
treatment means. In this situation, a mathematical transformation is commonly
performed on the entire data set to stabilize the variance of the residuals. For each shear
rate, a power transformation was applied to the data set. The lambda value which
normalized the variance of the residuals was different for each shear rate, and was
negative in each case. These values were computed by the Design Expert software, using
a Box-Cox plot. The adequacy of the generated models was very high with high R2,
adjusted R2 and adequate precision values generated for the relationship at each shear rate
(Table 3.3). ANOVA results showed that the curvature for the model was insignificant

(“Prob > F” less than 0.05), at each of the tested shear rates (see Appendix B). The

52

derived power regression equations describing the relationships between the true

viscosity and different variables (in terms of coded factors) at various shear rates were:

at 500 s"

(Viscosiiyr‘m = +0.063-5.5.1x10'3A -3.()3x10‘3B

(9)
+7.22x10'3C+ 5.69xl()'°AB-4.8lx10'3AC
at 1200 s"
(Vismsnyr‘ = +6.23 x 10' 3 - 1.04 x 10' 3 A - 0.39 x10'3 13 (10)
+1.31 x10'°C+ 0.39xllr'3AB-0.91x10'3AC
at 2000 s"
(Viscosity) 1-5 = +6.77 x 10' 4 - 1.69 x 10' 4 A - 0.36 x 10' 4 B (11)

+2.11x10'4C+1.169x10'4AB-1.63x10'4AC
where A, B, and C are the coded variables that represent wood flour content, acrylic

modifier content, and plasticizer content, respectively.

The above-described equations indicate that wood flour content, acrylic modifier
content and plasticizer content are the main effect factors, while wood flour
content/acrylic modifier content (interaction AB) and wood flour content/plasticizer
content (interaction AC) are the significant two-factor interaction terms. Since these are
power models with a negative lambda value, a positive algebraic sign of each factor in
Equations 9-1 1 denotes a decrease in viscosity, whereas a negative algebraic sign denotes

the opposite. As can be seen in the above equations, both wood flour content and acrylic

53

modifier content had a negative sign, indicating an increased effect on the viscosity of the
materials. In other terms, the true viscosity increased as the wood flour content or acrylic
modifier content increased in the composites. In contrast, the effect of plasticizer content
(factor C) was positive implying that the viscosity of the composites decreased with even
the low level of plasticizer content.

The two-level factorial design also revealed powerful interactions between wood
flour content and acrylic modifier content (interaction AB) as well as wood flour content
and plasticizer content (interaction AC) (Equations 9-11). The wood flour content/acrylic
modifier content interaction (interaction AB) had a negative sign, indicating a positive
effect on the true viscosity, while an opposite trend was observed for wood flour
content/plasticizer content interaction (interaction AC). As was previously mentioned, it
would be inappropriate to investigate these factors individually since the effect of one
factor will depend upon the level of the other factor in the model. The interaction graphs
based on Equations 9-11 are illustrated in Figure 3.3.

The viscosity of neat PVC matrix and PVC/wood-flour composites decreased
with increased plasticizer content, regardless of the presence of acrylic modifier and
extrusion shear rate investigated. However, the effect of plasticizer in decreasing the
viscosity was more pronounced in neat PVC than in the composite. This behavior was
expected due to the plasticization effect induced by the plasticizer. When a plasticizer is
incorporated into a PVC matrix, the polar groups of the plasticizer bond with the polymer
dipoles, and the non-polar parts act as shields between polymer dipoles, thus resulting in
less cohesion overall, with consequent increased freedom of molecular movement [27].

As a result, the chain flexibility or mobility increases, the matrix softens and flow

54

No acrylic modifier

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  
  

 

 

 

 

 

 

 

8 phr acrylic modifier
0.10 0.10
(a) to phrDEHP (b) +0 phrDEHP
0.09 i -l- 4 phrDEHP 0,09 4 +4 phrDEHP
V2 V2
5’ 0.08 ~ ? 0,03 4
4'4" F I
6'3 3:
e .:i
E' 1':
0.05 T I T T 0.05 T Y Y Y Y
-1.5 -1 -0.5 0 0.5 1 1.5 -1.5 -1 -0.5 0 0.5 1 1.5
Wood flour content Wood flour content
0.012 0012
d
(C) I ”NOE”, ( ) +4phrDB-P
—o— 0 phr new '4" 0 9'" DE”:
0.009 « 0.009 <
'1.- a;
I I
a a
v v
'r: 0.006 : \ '17 0-006 ‘
0.003 . . . 2 0.003 Y Y . . .
-15 -1 -0.5 o 0.5 1 1.5 -15 -1 -0-5 0 0-5 1 1.5
Wood flour content WOOd flour 00118..“
0.0014 , — 0.0014
(6) —s— o phr DEHP (0 +0 phr DEHP
"Ir-4 phrDEHP +4phrDEHP
I
2 0.0010 . 1: 0.0010 1
3' £3
a.
.7 2"
0.0002 7 r I 1 4 0.0002 1 Y T r Y
-1.5 -1 -0.5 0 0.5 1 1.5 -1.5 -1 -0.5 0 0.5 1 1.5
Wood flour content Wood flour content
Figure 3.3.

True viscosity at different shear rates: (a,b) at 500 s'1 (c,d) at 1200 s'1
and (e,f) at 2000 s'I as a function of the interaction between wood flour

and at 0 and 4 phr plasticizer content: (a,c,e) no acrylic modifier and
(b,d,f) 8 phr acrylic modifier.

55

properties are enhanced. As expected from the theory of particle suspension, the addition
of wood flour increased the viscosity of the polymer matrix, regardless of plasticizer
content and shear rate (Figure 3.3 a,c,e). It is believed that the presence of 50 phr wood
flour in the PVC matrix restricted molecular motion, imposing more resistance to flow
[9].

The viscosity of the neat PVC and the composite was a strong function of the
acrylic modifier content, independent of the plasticizer content and shear rate. However,
two distinct trends were observed: The addition of acrylic modifier significantly
increased the viscosity of the neat PVC [8], while it decreased the viscosity of the
composites (Figure 3.3 b,d,f). To firrther study this unusual behavior, a DMA analysis
was conducted on the neat PVC and composite with and without acrylic modifier. Table
3.4 shows the generated DMA data. Irrespective of the temperature the storage and loss
moduli of neat PVC increased with the addition of acrylic modifier (Table 3.4), implying
that the addition of acrylic modifier increased the melt elasticity and viscosity of the neat
PVC matrix. By contrast, the addition of acrylic modifier decreased the storage and loss
moduli of the composites. These results are in good agreement with those reported by
Guffey and Sabbagh [22], who showed that addition of CPE to PVC/wood-flour
composites reduced the viscosity. The probable reason for the acrylic modifiers opposite
effect observed on viscosity of PVC and composite may be due to its entanglement
effect. Acrylic modifiers have relatively high molecular weight (200,000 to 3 million)
compared to PVC (up to 85,000) [7]. When these compounds are melt mixed with PVC,
their long polymer chains entangle the short PVC chains, increasing the overall PVC melt

elasticity and viscosity (Figure 3.4) [7]. However, when wood flour is present in the

56

system, it may hinder the interaction between acrylic modifier and PVC, reducing the

entanglement effect.

57

 

 

 

mmd KN. w: 3M 36 3:6. $6 on: Uomm

Rd 3 .m 86 8.0 and 3.». Rd 8N Uooo

32¢ mm.m wmd N55 $6 34 Ed mod Doom.
2:35? 3.35? 3.25:. 3:63». 3.35;. 3.35: 0.235: 3.35:

maes— owamomm mmoa «aw—Sm .984 $985 981— «3.85

Aamwv Aux—6V mom—5.22.8“;

.3562: 31.8: 9 “MW”. .3512: Aamwv

5.. a a? £835 .. u 2?: a... a a? SE SE

 

3:32.25 a6_m-_.8><o>m .E 93
:32 .3 $2352 2:5er 3:352 awed ES Ana—=3: 3.3—MC 2:252 «MP—3m 2: no .3522 gamed. mo top—m— 1». «Sam.

58

 

 

Acrylic Modifier . Entanglements

n; W «a?

‘9

C; :3
I

*3 ' a:
r.

I
’

   
 
      

 

 

9

Vx

PVC Chains

 

.'

  

 

Figure 3.4. Entanglements of PVC Chains and Acrylic Modifier [7]

59

 

CONCLUSIONS

Using a factorial design approach, this study examined the effect of different
formulation variables on the viscoelastic properties of PVC/wood-flour composites
measured online on a conical twin-screw extrusion capillary rheometer. The following
conclusions can be drawn from the experimental results:

Wood flour content and acrylic modifier content were the two important variables
affecting the die swell ratio, whereas the addition of a low level of plasticizer had no
effect. The die swell increased with increased acrylic modifier content, regardless of the
addition level of wood flour. This was a probable result of increased melt elasticity of
the polymer. However, the die swell ratio was considerably reduced by adding wood
flour into the PVC matrix as a result of reduced normal stresses and lower coefficient of
thermal expansion.

The true viscosity of neat PVC and PVC/wood-flour composites decreased with
the plasticizer content, irrespective of the acrylic modifier content due, to the
plasticization effect of the plasticizer. However, the addition of acrylic modifier
significantly increased the viscosity of unfilled PVC due to increased melt elasticity.
Unlike for neat PVC, the true viscosity decreased as the acrylic modifier content

increased in the composites.

6O

10.

11.

REFERENCES

Markarion, J ., "Antimicrobial plastics additives: trends and latest developments in
North America", Plastics, Additives and Compounding, 4(12): 18-21 (2002).

Matuana, L.M. and Mengeloglu, F., "Manufacture of rigid PVC/wood-flour
composite foams using moisture contained in wood as foaming agent", Journal of
Vinyl & Additive Technology, 8(4): 264-270 (2002).

Li, Q. and Matuana, L.M., "Foam extrusion of high density polyethylene/wood
flour composites using chemical foaming agents", Journal of Applied Polymer

Science, 88(14): 3139-3150 (2003).

Matuana, L.M. and Mengeloglu, F., "Microcellular foaming of impact-modified
rigid PVC/wood-flour composites", Journal of Vinyl & Additive Technology, 7(2):
67-75 (2001).

Matuana-Malanda, L., Park, C.B., and Balatinecz, J.J., "Characterization of
microcellular foamed PVC/cellulosic—fiber composites", Journal of Cellular
Plastics, 32(5): 449-469 (1996).

Mengeloglu, F. and Matuana, L.M., "Foaming of rigid PVC/wood flour
composites through a continuous extrusion process", Journal of Vinyl & Additive
Technology, 7(3): 142-148 (2001).

Disson, J.-P. and Girois, 3., "Acrylic process aids for PVC: From theoretical
concepts to practical use", Journal of Vinyl and Additive Technology, 9(4): 177-
187(2003)

Kitai, K., Holsopple, P., and Okano, K., "Selecting the proper process aid for the
application", Journal of Vinyl and Additive Technology, 14(4): 211-217 (1992).

Matuana, L.M., Park, C.B., and Balatinecz, J.J., "The effect of low levels of
plasticizer on the rheological and mechanical properties of polyvinyl

chloride/newsprint-fiber composites", Journal of Vinyl & Additive Technology,
3(4): 265-273 (1997).

Collins, EA. and Krier, C.A., "Poly(vinyl chloride) melt rheology and flow
activation energy", Transactions of the Society of Rheology, 11(2): 225-42 (1967).

Sieglaff, C.L., "Rheological behavior of poly(vinyl chloride) mixtures. I. Viscous
behavior", Polymer Engineering & Science, 9(2): 81-85 (1969).

61

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

Glomsaker, T., Hinrichsen, EL, and Thorsteinsen, P., "In-line versus off-line
rheological measurements on S-PVC formulations releant for flow in extrusion
dies", Polymer Engineering & Science, 41(12): 2231-2248 (2001).

Berard, M.T., "Viscosity effects in rigid PVC", Journal of Vinyl and Additive
Technology, 6(3): 140-145 (2000).

Utracki, L.A., Bakerdjian, Z., and Kamal, M.R., "Flow behavior of semirigid and

rigid poly(vinyl chloride) formations", Transactions of the Society of Rheology,
19(2): 173—99 (1975).

Shah, P.L., "Significance of capillary rheology testing temperture for predicting
PVC extrudability", Polymer Engineering and Science, 14(11): 773-7 (1974).

Collins, E.A., Metzger, AP, and Furgason, R.R., "Relation between Brabender

and capillary rheometer thermal stability measurements for PVC compounds",
Polymer Engineering and Science, 16(4): 240-5 (1976).

George, J ., Janardhan, R., Anand, J.S., Bhagawan, 8.8., and Thomas, 8., "Melt
rheological behaviour of short pineapple fibre reinforced low density
polyethylene composites", Polymer, 37(24): 5421-5431 (1996).

Xiao, K. and Tzoganakis, C., "Rheological properties and their influence on
extrusion characteristics of HDPE-wood composite resins", Annual Technical
Conference - Society of Plastics Engineers, 60th(Vol. 1): 252-256 (2002).

Marcovich, N.E., Reboredo, M.M., Kenny, J ., and Aranguren, M.I., "Rheology of
particle suspensions in viscoelastic media. Wood flour-polypropylene melt",
Rheology Acta, 43: 293 (2004).

Maiti, SN. and Hassan, M.R., "Melt rheological properties of polypropylene-
wood flour composites", Journal of Applied Polymer Science, 37(7): 2019-2032
(1989)

Goettler, L.A., "The extrusion and performance of plasticized poly(vinyl chloride)
hose reinforced with short cellulose fibers", Polymer Composites, 4(4): 249-255
(1983).

Guffey, V0. and Sabbagh, A.B., "PVC/wood flour composites compatibilized
with chlorinated polyethylene", Annual Technical Conference - Society of Plastics
Engineers, 60th(Vol. 3): 3315-3319 (2002).

62

23.

24.

25.

26.

27.

Montgomery, D.C., Design and analysis of experiments. 5th ed, New York: John
Wiley. xii, 684 p. (2001).

Middleman, S., The flow of high polymers; continuum and molecular rheology,
New York,: Interscience Publishers. ix, 246 p. (1968).

Dealy, J .M. and Wissbrun, K.F., Melt rheology and its role in plastics processing
.' theory and applications, New York: Van Nostrand Reinhold. xix, 665 p. (1990).

Patterson, J. and Szamborski, 6., "Expanding PVC as a building material",
Journal of Vinyl & Additive Technology, 1(3): 148-54 (1995).

Sears, J .K. and Darby, J .R., The technology of plasticizers, New York: Wiley. xi,
1166 p. (1982).

63

CHAPTER 4
CONTINUOUS EXTRUSION OF MICROCELLULAR FOAMED

PVC AND ITS WOOD FLOUR COMPOSITES USING
SUPERCRITICAL CO;

64

ABSTRACT

The aim of this study was to produce microcellular foamed PVC and its wood
flour composites in a continuous extrusion process using supercritical CD; as a blowing
agent. An extrusion foaming setup was built to accomplish this goal. Various material
composition and processing parameters were studied to determine their effects on the
morphology of the foamed samples. Although foamed PVC samples could be produced
from this setup, their processing-property relationships could not be determined due to a

lack of steady-state conditions during the foaming process.

Keywords: Microcellular foaming, Polyvinyl chloride, Supercritical fluid, Extrusion

65

INTRODUCTION

Wood plastic composites (WPCS) represent a unique and growing share of the
building products market. With advantages including durability, visual appeal,
customizability, ease of installation and cost effectiveness compared to solid wood and
other traditional building materials, WPCS are gaining market share in a variety of
applications. Several companies have launched innovative lines of exotic-looking decking,
railing systems, privacy fencing and complimentary accessories in recent years that have
improved consumer acceptance of, and desire for, WPCS. If these trends continue, the
market can be expected to grow and develop in new directions. For example, the market for
extruded WPCs was approximately $1 billion in 2005 and the market is expected to grow at
14% through 2010 [1, 2].

WPCS use wood flour or fiber as a filler or reinforcing agent in the plastic matrix.
Wood flour has several advantages over conventional fillers (talc and calcium carbonate),
advantages include low density, low cost and less abrasiveness to processing equipment.
A variety of polymers have been used in manufacturing WPCS including polyethylene
(PE), polypropylene (PP), polyvinyl chloride (PVC), polystyrene (PS) and acrylonitrile
butadiene styrene (ABS). These polymers may be virgin or recycled, and in either case,
may be filled with wood that has been reclaimed from other uses.

PVC is one of the most-used polymers used in manufacturing WPCs on a volume
basis, although the market is led by polyethylene. PVC wood composites offer several
advantages over polyolefin-based wood composites such as increased weatherability,
high stiffiiess, ease of customizing the final product properties by changes in formulation,

and good paintability of the final product. The specific gravities of polyolefins and PVC

66

are higher than that of solid wood. When combined into WPCS, the resulting composite
materials have higher specific gravity than solid lumber. Compared to polyolefin-based
WPCS, the PVC wood composites are even denser, because PVC is denser than
polyolefins. The high specific gravity of PVC wood composites makes them heavier for
shipping, handling and end use applications. Addition of wood to the PVC also
significantly reduces the impact strength of the PVC matrix. These drawbacks have been
addressed by producing microcellular-foamed structures in WPCs. This reduces the
density of the composites. Additional advantages of microcellular foamed WPCS include
reduced material costs and increased mechanical properties, such as impact strength [3,
4]. The combination of property enhancement and cost reduction makes microcellular
foamed WPCs attractive for use in a variety of innovative applications, not limited to
building products.

The simplest way to introduce a microcellular structure into samples is to use a
batch foaming process. Polymers of differing morphologies such as HDPE [5, 6], PP [6],
PVC [7], PS [8], PC [9], etc., have been foamed using this technique. Most of the
research on microcellular foaming of PVC and its composites has used the batch process.
The relationships between void fraction and mechanical properties of batch foamed PVC
and its composites have been investigated with varying results [10, 11]. An in-depth
study is still needed to evaluate the dependence of mechanical properties on the
characteristics of microcellular foamed PVC.

In spite of the ease and simplicity of batch foaming, this process is not likely to be
implemented in the industrial production of foams because it is not cost-effective. The

microcellular batch foaming process is time consuming due to the multiple steps involved

67

in the production of foamed samples - particularly the sample saturation step, which can
last from several hours to days. To overcome these drawbacks, a continuous process of
manufacturing microcellular foam using an extrusion technique was proposed by several
researchers [12-15].

Park et al. [16] studied the effects of different processing parameters on the
morphology of high impact polystyrene (HIPS) foams using an extrusion technique.
They found that melt temperature had a significant effect on cell coalescence, which
could be prevented by selecting the correct melt and die temperatures. Reducing the die
temperature from 170 to 135 0C increased the volume expansion ratio of the foam,
mainly because lower temperatures reduced the gas loss. Further decrease in temperature
had a negative effect on the volume expansion ratio of the foamed HIPS samples because
of the increased stiffness of the polymer matrix.

Zhang et al. [17] used the single and tandem extrusion processes to foam
microcellular HDPE/wood fiber composites using CD; as a blowing agent. They showed
that increasing wood fiber content deteriorated the cell morphology in terms of both cell
size and number. Similarly, when the C02 content was increased above 1 wt. % there no
improvement in the cell structure and morphology was observed. It was also shown that
the addition of 5 wt. % coupling agent dramatically improved the cell structure in the
composites. This was attributed to improved bonding between the wood fiber and HDPE
matrix due to the presence of the coupling agent.

The continuous extrusion process has been successfully used for several polymers
and their composites, but to date there has been no available literature on using this

process for PVC and its wood composites using online injection of supercritical fluid as

68

the blowing agent. The limited research done on extrusion foaming of PVC and its
composites has been mainly focused on chemical blowing agents, though the use of
moisture as a foaming agent has also been reported in the literature [18]. Chemical
blowing agents require extra care during dry blending to avoid migration, which results in
density variations in the final product. Apart from this, chemical blowing agents have
several other disadvantages including difficulties in adjusting online foam densities [19].
Physical blowing agents such as carbon dioxide (C02), nitrogen (N2) and argon (Ar) are
economical and eco—friendly, which makes them a favorable alternative to chemical
blowing agents.

One challenge to producing PVC composite foams is overcoming the increase in
the viscosity of the PVC matrix when wood flour is added. Increased viscosity hinders
the bubble formation and growth and generally results in poor foam quality in terms of
cell size and density [20, 21]. Our prior work (chapter 3) showed that the melt viscosity
of PVC/wood-flour composites could be successfully altered by the addition of acrylic
foam modifier [22].

Applying the continuous extrusion process technique to PVC and its wood flour
composites, along with the use of supercritical C02 as a blowing agent, would result in an
efficient, environmentally-friendly and cost-effective process. Thus, the goal of this
study was to manufacture microcellular foamed PVC and its composites in a continuous

extrusion process using supercritical fluids.

69

EXPERIMENTATION

Materials:

PVC, K value-57 (OxyVinyls) was used as a matrix in this study (Table 4.1). Tin
stabilizer (PlastiStab 2808) from OMG Americas was used as a heat stabilizer. Calcium
stearate (Synpro, Ferro Corp.) and paraffin wax (Gulf Wax) were used as lubricants.
Different acrylic processing aids (Paraloid K-125, Paraloid K-175, Rohm and Haas Co.)
were also used in the formulations. Acrylic foam modifier (Paraloid K-400, Rohm and
Haas Co.) was used to facilitate the foaming process. Wood flour from hardwood maple

(American Wood Fibers) was used as filler. CO2 was used as a physical blowing agent.

Continuous Foaming Process:

The ingredients were weighed at the addition levels described in Table 4.1 and
were mixed in a high intensity batch mixer (Papenmeier, TGAHK20) for 10 min at room
temperature. Before compounding, wood flour was dried in an oven for 48 hrs at 105°C
to a moisture content of less than 1%. The compounded mixture was then fed through a
19 mm single screw extruder with an L/D ratio of 25:1 (C. W. Brabender Instruments
Inc.) (Figure 4.1). Supercritical CO2 gas was injected at different concentrations into the
extruder under high pressure between the second and third zones of the barrel. The
concentration of the gas was precisely controlled by a positive displacement syringe
pump (ISCO 260D). A static mixer (Omega F X 844418) was attached at the end of the
extruder to ensure homogenization of the solution of polymer melt and the blowing gas,

followed by a custom designed nozzle die (Figure 4.1).

70

Table 4.1. Formulations Used in Manufacturing Microcellular F oamed PVC and its
Wood-Flour Composites

 

 

Ingredients Concentration (phr)
PVC K-57 100
Tin stabilizer 2
Calcium stearate 1.5
Paraffin wax 2
Paraloid K-120 0 — 2
Paraloid K-175 O — 2
Paraloid K-400 0 — 8
Wood flour 0 —- 30

 

71

 

Valves

4.

Syringe Pump
Gas Pressure

Transducer :1

 

 

 

Hopper
\ Melt Temperature

 

 

 

 
  
   

Motor

        

’

\\l \\\\\\\\l \\\\]\]_\ (as .m.

”Mg—- —-1—-

  

 

 

 

 
  

Static Mixer T

Mixing Zone Melt pressure

Water in Water out

Figure 4.1. Schematic of Continuous Microcellular Foaming Process

72

CO2 Tank

Nucleation Die

Thermodynamics of Foaming:

The concentration of gas and the time required to induce foaming in a polymer
depend on the solubility and diffusion coefficient of the gas in the polymer matrix. Both
the solubility and diffusivity of the gas vary, depending on the processing parameters
such as temperature and pressure. The diffusion coefficient increases with temperature,
which can be expressed by the following Arrhenius type equation [23]:

D = D0 exp(-AED /RT)
where D is the diffusion coefficient, Do is the diffusion coefficient constant, ED is the

activation energy, R is the universal gas constant and T is the temperature.

Muth et al. [24] showed that increasing the temperature and pressure of the
system resulted in higher diffusivity of CO2 gas in PVC polymer. Raising the
temperature of PVC from 40 to 70 0C increased the diffusivity of CO2 gas from 0.7 x 10'
H mZ/s to 2.51 x 10'H mZ/s at 20 MPa pressure. As a result, saturation time for a sample
of thickness 2mm is reduced from 40 hours at 40 °C to 11 hours at 70 OC, and could be
further reduced to several minutes at elevated temperatures. The time required by the gas

to diffuse into the sample can be calculated by the following equation [23]:
z, =13 /D
where td is the diffusion time, 1d is the diffusion thickness/distance and D is the diffusion
coefficient.
The higher temperatures and pressures used in the extrusion foaming process are

expected to lower the time required for sample saturation. It will be important to select

the proper processing speed, which will provide long enough residence time during

73

extrusion to ensure the complete saturation of the polymer with blowing gas to form a

single phase solution.

For the extrusion foaming process to be successful, an adequate amount of
blowing gas must be dissolved in the polymer. Gas solubility in a polymer increases with
increasing pressure, but decreases with increasing temperature. The relationship between
the solubility or concentration of the gas and pressure can be expressed by Henry’s Law:

C = H*P
where C is the gas saturation concentration, H is the Henry’s Law Constant and P is the
saturation pressure.

From Henry’s law it is clear that high pressure must be maintained during
processing to increase the solubility of the gas in the polymer. This becomes even more
critical at high temperatures. Muth et al. also demonstrated that at 20 MPa pressure, an
increase in temperature from 40 to 70 oC reduced the CO2 solubility in PVC by nearly
15%. Similarly, an increase in temperature from 100 to 200 oC reduced the solubility of
CO2 in polystyrene from 7.17 wt. % to only 4 wt. % at 12.6 MPa pressure [25]. There is
no solubility data available in the literature for PVC at elevated temperature and pressure.
However, based on the lower temperature data for PVC, the solubility of CO2 in PVC at
extrusion temperatures is expected to be reduced even further. High pressure within the
extruder barrel will be required to maintain adequate saturation of the polymer with

blowing gas for foaming the PVC.

74

Physical Characterization of Foams:
Microcellular foam can be characterized by the proportion of voids in the foamed
sample, which is known as the void fraction. The void fraction (Vf) is an indication of

the foam density. It was calculated using the following equation:

Vf- pf

p

where pf and p are the densities of the foamed and unfoamed sample, respectively.

The density of the samples was determined by the water displacement method as per

ASTM standard D-792 and was calculated by the following formula:

 

Density (g/cm3) = 0.9975(3)

W

where Ma and Mw are the weights of samples measured at room temperature in air and

water, respectively.

SEM Analysis:

Scanning electron microscopy (SEM) was used to study the morphological
characteristics of the extruded foamed samples. This analysis shows the cell density and
cell size in the foamed samples, which are major criteria in characterizing microcellular

foam. The cell density (Nf) was determined by the following formula:

M2 3/2 1
" 1 1
A l-V/

 

 

75

where n is the number of cells in the area A, and M is the magnification factor of the
micrograph.

SEM analysis was performed on fracture surfaces of cryogenically fractured foam
samples. A JEOL JSM-6400 scanning electron microscope equipped with a LaB6 gun
was used to collect the SEM images. An accelerating voltage of 12KV was used. The

samples were gold coated in a sputter coater before collecting the images.

76

RESULTS AND DISCUSSION

Experimental Setup:

Extruding continuous, perfect foam requires constant feeding, conveying, melting
and mixing of the polymeric material before it reaches the injection stage. To accomplish
this, a 19 mm single screw extruder with a 25:1 L/D ratio (C.W. Brabender) was used
(Figure 4.1). The extruder had three separate heating zones to effectively control the
temperature of the polymer melt. After melting the polymer material, the blowing agent
was continuously injected at a constant flow rate and pressure to produce a uniformly
foamed product. A positive displacement pump (ISCO D260) with a custom designed
injection port and flow restrictor (Mott Corporation) was used to achieve the constant
flow rate and gas pressure. To ensure adequate mixing of the polymer and CO2 after
injecting the blowing agent, the injection port was placed between the second and third
zones of the extruder barrel (Figure 4.1). The injection port consisted of a three inch
blank melt bolt with a hole drilled through it to permit the passage of gas. To
successfully inject the gas into the extruder, the gas pressure had to be higher than the
actual barrel pressure. Otherwise, the polymer would flow towards the injection port and
block it, preventing the blowing agent from flowing through the extruder. Supercritical
CO2 is a highly compressible fluid; this results in variation of the flow rate due to
changes in the barrel pressure. This can be avoided by injecting the blowing agent at
very high pressure, compared to the pressure in the extruder barrel. This pressure
gradient was maintained with a flow restrictor, which is nothing but a fine metal
membrane. The flow restrictor created resistance for the gas flow and thus maintained its

high pressure. It also prevented the polymer from entering the injection port and

77

blocking it. The flow restrictor was machined into the very bottom of the injection port
(blank melt bolt). Once the gas was effectively injected into the extruder, it had to be
mixed with the polymer melt. A specially designed mixing area on the screw was used to
achieve the mixing. This section of the screw provided extra shear, resulting in better
mixing. To further increase the mixing and form a homogeneous solution of polymer and
gas, a 6-element static mixer (Omega FMX844IS) was attached to the extruder (Figure
4.1). The element of the static mixer had a left and right helix, ensuring a vigorous and
homogenous mixing of polymer melt and gas. To keep the blowing agent soluble in the
polymer melt it had to be maintained at high pressure. A custom designed nozzle die was
connected at the end of the static mixer (Figure 4.1). A 1 mm diameter nozzle was used
to achieve a high pressure drop at the die opening, which results in thermodynamic
instability by reducing the gas solubility in the polymer/gas solution. This instability
created the phase separation and initiated the nucleation of numerous cells in the polymer
matrix, which was then immediately cooled to preserve the cell structure. This

experimental setup was based on the process described by Park et a1. [18].

Microcellular Foaming of PVC:

As described above, the setup of an extrusion system for continuous foaming of
polymers and composites using supercritical fluid as a blowing agent is very complex and
involves many variables. To ensure that the experimental setup was adequate to foam
polymeric materials, HDPE was used as a baseline. Other researchers have demonstrated
the continuous foaming of HDPE in similar extrusion setups [19, 25]. Conditions

described in these studies were used as the basis for foaming HDPE. These included die

78

temperatures ranging from 140 — 175 °C, extrusion speeds of 20-40 rpm and gas flow
rates between 0.1 — 2 ml/min.

Successful microcellular foamed samples of HDPE were obtained under various
combinations of temperature, gas flow rate and extrusion speed. Steady state conditions
of constant pressure, temperature and gas flow rate were attained over this range. Steady
state conditions are crucial for foam characterization, since, as mentioned previously,
CO2 gas is highly compressible. The density of CO2 gas is directly proportional to the
pressure and inversely proportional to the temperature. Thus, any changes in pressure
and temperature will change the net mass flow rate of the gas. This makes it difficult to
characterize the processing-property relationships of the foamed samples, as the exact
amount of CO2 used to manufacture the samples is not known. SEM micrographs of
representative samples of microcellular foamed HDPE were taken to visualize the
morphology of the samples (Figure 4.2). The top image (Figure 4.2a) shows
microcellular sized bubbles that are distributed fairly evenly throughout the sample. This
implies that the designed setup can produce good quality microcellular foams. Figure
4.2b shows the cell wall area at higher magnification. This figure provides evidence of
even smaller bubbles that are very evenly distributed throughout the cell wall area. These
bubbles are less than a micron in size. HDPE foamed very well in the extrusion process
using CO2 as a blowing agent, which suggested that the extrusion foaming setup was
properly designed to achieve continuous microcellular foaming of polymers using a

supercritical fluid blowing agent.

79

 

Figure 4.2. SEM Micrographs of HDPE at (a) 100X and (b) 10,000X

80

Unlike HDPE, PVC was not easily foamed using the continuous extrusion process
with C02. The injection of supercritical fluid CO2 into the extruder did not provide even
gas pressure in the extruder during foaming. It was not possible to reach and maintain
steady state conditions where the pressure was held constant and the material would exit
the die as a foam. Instead, the pressure fluctuated greatly throughout the experiments,
resulting in some partially foamed samples that were rapidly ejected from the die under
high pressure. The various combinations of barrel temperature, die temperature and
extrusion speed that were tested for foaming PVC are listed in Table 4.2. The gas content
was varied from 0.5 — 5 wt. %. The inability to reach and maintain steady state
conditions may have been due to the high melt viscosity of PVC melt compared to
HDPE, which could result in very high pressure drops in the static mixer and die. This
pressure drop reduces the solubility of the gas in the polymer and foams the polymer
matrix well before the die exit. The released gas inside the extruder system expands due
to the high temperature and lack of high pressure, creating a huge empty pocket in the
system and thus breaking the continuous extrusion of polymer. This released gas exits
the die intermittently, shooting the polymer from the die at high pressure. Attempts were
made to suppress this behavior by changing the viscosity of the PVC melt.

Our previous research [24] (Chapter 3) showed that additives such as foam
modifier, plasticizers and filler have a significant effect on the melt viscosity of the PVC
melt. Based on that study, the batch composition of the PVC was altered in favor of
lower melt viscosity by adding 4 and 10 phr of DEHP plasticizer. Various combinations
of extruder screw speed, barrel temperature and die temperature (listed in Table 4.2) were

again tested with these formulations to determine whether lower melt viscosity might

81

Table 4.2. Different Processing Conditions used to Manufacture Microcellular
Foamed PVC and Its Composites

 

 

Zone 1 Zone 2 Zone 3 Die Screw Speed
(° C) (° C) (° C) (o C) (RPM)
160 180 175 150-180 10-30
160 185 180 150-180 10-30
180 190 185 150-180 10-30
175 190 180 150-180 10-25
175 180 175 160-175 10-20

 

82

lead to steady state conditions and improved foaming. However, none of these tests were
successful, as steady state conditions were never achieved with PVC. Other polymers
such as polystyrene (both regular and high impact) and polyethylene of various melt flow
rates were also tested in this extruder setup to see whether these variables would have an
effect on foaming. Surprisingly, all polymers tested, other than PVC, were able to be
foamed successfully in this setup. The success obtained with other polymers suggested
that the PVC polymer itself was likely the obstacle to foaming. The complex
processability of PVC, including difficulty in controlling its melt viscosity, was one
probable reason for the difficulties in foaming. Another possible problem might have
been insufficient mixing of the polymer melt and gas due to the small mixing zone in the
extruder. This could be eliminated by using a longer L/D ratio extruder. Lack of a
proper flow restrictor, which could maintain a high enough pressure difference between
the barrel and gas pressure, may be another reason why the PVC did not consistently
foam.

Some foamed samples were collected from the extruder, but these were not
obtained at steady state conditions. SEM micrographs were collected for foamed samples
produced with varying content of CO2 (Figures 4.3 and 4.4). Since the pressure and mass
flow rate were not stable at any time during the run, the effects of various processing
conditions and material composition could not be systematically compared.

Any trends observed during non-steady state conditions were not necessarily
reflective of actual behavior of the foam. An overall View of the foamed PVC sample at

3 wt. % CO2 content (Figure 4.3a) showed randomly distributed large and small bubbles

83

 

Figure 4.3. SEM Micrographs of PVC foamed at 3 wt. % C02 (a) 20X and (b) 50X

84

throughout the sample. This indicated that the polymer and gas were not mixed properly
and a homogeneous solution was not formed, which resulted in uneven foaming of the
sample. A magnified view of the same sample (Figure 4.3b) illustrates a broken cell
structure that was consistent throughout the sample. This can result either from high
foaming temperature or from premature foaming inside the machine before exiting the
die. To study whether this was a result of temperature, a wide range of die temperatures
were studied, ranging from 135 to 175 °C. Samples collected at various temperatures
over this range displayed similar cell morphology. Thus, we concluded that the pressure
inside the machine was not high enough to maintain the solubility of the C02 gas in the
polymer melt, resulting in premature foaming of the samples. High pressure in the
system could not be maintained due to the excessive pressure drop in the static mixer and
die, resulting from the high viscosity of the PVC melt. In the SEM micrographs of
samples foamed with 2 and 1 wt. % of C02, respectively (Figure 4.4a and 4.4b), larger
bubbles were formed at the lower mass flow rate of C02. None of the SEM images show
bubbles that were uniform in size. While some of the bubbles appear to be microcellular,
many are larger than that.

Since steady state conditions were not achieved, it was not possible to assess the
effects of processing conditions or material composition on the foam structure of the
PVC with any certainty. For the several reasons discussed above, including the lack of
sufficient mixing after the gas was introduced into the extruder, we were unable to attain

the initial goals of this experiment.

85

 

Figure 4.4. SEM Micrographs of PVC foamed at (a) 2 wt. % CO2 and (b) 1 wt. %

C02

86

CONCLUSIONS

The goal of this study was to produce microcellular foamed PVC and its wood
flour composites through a continuous extrusion process using supercritical C02 as the
blowing agent. A continuous extrusion setup for microcellular foaming was developed in
this study. Polymers such as HDPE and PS were successfully microcellular foamed
under steady-state conditions with these this setup. However, when PVC was foamed,
steady-state conditions could not be achieved. This may be due to insufficient mixing of
polymer and blowing gas and to a lack of high enough pressure to keep the gas soluble in
the polymer. SEM images revealed uneven cell size and extreme cell wall breakage
within the PVC foams. This confirmed that there was a lack of proper mixing and
premature foaming in the extruder, due to the lack of high pressure needed to maintain
the solubility of C02 gas in the polymer melt. Attempts were made to control this
behavior by testing various processing conditions and material compositions. However,
none of these attempts were able to produce good quality microcellular foamed samples

under steady-state conditions.

87

10.

REFERENCES

Smith, PM. and Wolcott, M.P., "Opportunities for Wood/Natural Fiber-Plastic
Composites in Residential and Industrial Applications", Forest Products Journal,
56(3): 4-11 (2006).

Morton, J., Quarrnley, J., and Rossi, L. "Natural and Woodfiber Composites in
North America and Europe". in 7th International Conference on Woodfiber-Plastic
Composites (and other natural fibers). 2003, Madison, WI: Forest Product Society.

Mengeloglu, F. and Matuana, L.M., "Foaming of rigid PVC/wood flour composites
through a continuous extrusion process", Journal of Vinyl & Additive T echnologz,
7(3): 142-148 (2001).

Matuana-Malanda, L., Park, C.B., and Balatinecz, J.J., "Characterization of

microcellular foamed PVC/cellulosic-fiber composites", Journal of Cellular
Plastics, 32(5): 449-469 (1996).

Doroudiani, S., Park, C.B., Kortschot, M.T., and Cheung, L.K., "Effect of
morphology on microcellular foaming of semi-crystalline polymers", Annual
Technical Conference - Society of Plastics Engineers, 53(Vol. 2): 2183-8 (1995).

Doroudiani, S., Park, C.B., and Kortschot, M.T., "Effect of the crystallinity and
morphology on the microcellular foam structure of semicrystalline polymers",
Polymer Engineering and Science, 36(21): 2645-2662 (1996).

Kumar, V. and Weller, J.E., "A process to produce microcellular PVC",
International Polymer Processing, 8(1): 73-80 (1993).

Kweeder, J.A., Ramesh, N.S., Campbell, G.A., and Rasmussen, D.H., "The
nucleation of microcellular polystyrene foam", Annual Technical Conference -
Society of Plastics Engineers, 49: 1398-400 (1991).

Kumar, V. and Weller, J.E., "Microcellular polycarbonate - Part 1: Experiments on
bubble nucleation and growth", Annual Technical Conference - Society of Plastics
Engineers, 49: 1401-5 (1991).

Matuana, L.M., Park, C.B., and Balatinecz, J.J., "Structures and mechanical
properties of microcellular foamed polyvinyl chloride", Cellular Polymers, 17(1):
1-16(1998).

88

ll.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

Juntunen, R.P., Kumar, V., Weller, J.E., and Bezubic, W.P., "Impact strength of
high density microcellular poly(vinyl chloride) foams", Journal of Vinyl & Additive
Technology, 6(2): 93-99 (2000).

Baldwin, D.F., Park, C.B., and Sub, N.P., "A microcellular processing study of
poly(ethylene terephthalate) in the amorphous and semicrystalline states. Part II.
Cell growth and process design", Polymer Engineering and Science, 36(11): 1446-
1453(1996)

Baldwin, D.F., Park, C.B., and Sub, N.P., "An extrusion system for the processing
of microcellular polymer sheets: shaping and cell growth control", Polymer
Engineering and Science, 36(10): 1425-1435 (1996).

Behravesh, A.H., Park, C.B., and Venter, R.D., "Extrusion of low-density
microcellular HIPS foams using C02", Aw (American Society of Mechanical
Engineers), 76(Cellular and Microcellular Materials): 47-63 (1996).

Park, C.B., Pan, M., and Yeung, A.K. "Microcellular processing of low-density PS
foams in extrusion". in Cellular Polymers IV, International Conference, 4th,
Shrewsbury, UK, June 5-6, 1997. 1997.

Park, C.B., Behravesh, A.H., and Venter, R.D., "Low density microcellular foam
processing in extrusion using C02", Polymer Engineering and Science, 38(11):
1812-1823 (1998).

Zhang, H., Rizvi, G.M., and Park, C.B., "Development of an extrusion system for
producing fine-celled HDPE/wood-fiber composite foams using C02 as a blowing
agent", Advances in Polymer Technology, 23(4): 263-276 (2004).

Matuana, L.M. and Mengeloglu, F., "Manufacture of rigid PVC/wood-flour
composite foams using moisture contained in wood as foaming agent", Journal of
Vinyl & Additive Technology, 8(4): 264-270 (2002).

Dey, S.K., Jacob, C., and Xanthos, M., "Inert-gas extrusion of rigid PVC foam",
Journal of Vinyl & Additive Technology, 2(1): 48-52 (1996).

Lee, S.T., Foam extrusion .' principles and practice, Lancaster: Technomic Pub.
Co. xviii, 344 p. (2000).

Okamoto, K.T., Microcellular processing. lst ed, Cincinnati, OH: Hanser Gardner
Publishers. xiii, 186 p. (2003).

89

22.

23.

24.

25.

26.

Shah, BL. and Matuana, L.M., "0nline measurement of rheological properties of
PVC/wood-flour composites", Journal of Vinyl & Additive Technology, 10(3): 121-
128 (2004).

Park, CB. and P., S.N., "Rapid polymer/gas solution formation for continous
production of microcellular plastics", Journal of Manufacturing Science and
Engineering, 118: 639-645 (1996).

Muth, 0., Hirth, T., and Vogel, H., "Investigation of sorption and diffusion of

supercritical carbon dioxide into poly(vinyl chloride)", The Journal of Supercritical
Fluids, 19(3): 299-306 (2001).

Sato, Y., Takikawa, T., Takishima, S., and Masuoka, H., "Solubilities and diffusion
coefficients of carbon dioxide in poly(vinyl acetate) and polystyrene", The Journal
of Supercritical Fluids, 19(2): 187-198 (2001).

Anderson, J ., Blizard, K., Okamoto, K., and Straff, R. "Microcellular HDPE foam
for blow molding". in Foams ’99, International Conference on Thermoplastic Foam
1999, Parsippany, NJ, United States.

90

CHAPTER 5

NOVEL COUPLING AGENTS FOR PVC/WOOD-FLOUR
COMPOSITES’

This chapter was expanded from the paper published in the Journal of Vinyl and Additive
Technology, 11 (4), 160-165 (2005).

mThis paper was the winner of the 2005 Best Student Paper Award presented by the SPE
Vinyl Division

91

ABSTRACT

Effective interfacial adhesion between wood fibers and plastics is crucial for both
the processing and ultimate performance of wood-plastic composites. Coupling agents
are added to wood-plastic composites to promote adhesion between the hydrophilic wood
surface and hydrophobic polymer matrix, but to date no coupling agent has been reported
for PVC/wood-fiber composites that significantly improved their performance and was
also cost-effective. This paper presents the results of a study using chitin and chitosan,
two natural polymers, as novel coupling agents for PVC/wood-flour composites.
Addition of chitin and chitosan coupling agents to PVC/wood-flour composites increased
their flexural strength by approximately 20%, their flexural modulus by approximately
16%, and their storage modulus by approximately 33-74%, compared to PVC/wood-flour
composite without the coupling agent. Significant improvement in the composite

performance was attained with 0.5 wt% chitosan and while 6.67 wt% chitin.

Keywords: PVC, PVC/wood-flour composites, coupling agent, chitin, chitosan,

dynamic mechanical analysis, flexural properties.

92

INTRODUCTION

Wood-plastic composites (WPCS) have emerged as an important family of
engineering materials. They are partially replacing solid pressure-treated wood and other
materials in a variety of applications [1]. Although WPCs are superior to the neat
polymers in terms of material cost and stiffness, their strength performance (tensile,
flexural, and impact) is generally lower compared to the unfilled polymers [2,3]. The
decreased strength is likely a result Of the natural incompatibility of the hydrophilic wood
fibers (high surface tension) and the hydrophobic polymer matrix (low surface tension) [2-
5]. Phase incompatibility yields very weak interactions and thus a weak interface (poor

interfacial adhesion) between the fiber and the matrix.

One approach to designing WPCS is to modify the wood fiber surface with
coupling agents to improve the strength [6]. Coupling agents convert the hydrophilic
surface of wood fibers to a more hydrophobic surface. Thus, the surface tension of the
wood fibers is reduced and approaches that of the molten polymer. As a result, wetting and
adhesion are improved via mechanisms such as diffusion and mechanical interlocking

between treated fibers and the polymer matrix [6].

Due to its strong effect in altering the hydrophilic surface of wood fibers to a more
hydrophobic one, maleic anhydride functionalized polyolefin is commonly used as a
coupling agent for polyolefin/wood-fiber composites [4,6]. Similarly, to enhance the
interfacial adhesion between wood fibers and a PVC matrix, the second largest plastic
used to manufacture WPCS, several investigators have assessed the effects of various

fiber treatments, including different types of isocyanates, maleic anhydride, silanes, etc.

93

as coupling agents [7-9]. Most mechanical properties Of the composites were improved
by these chemical treatments, compared to those of composites with non-treated fibers.
However, the properties of the composites were inferior tO those of unfilled PVC,
suggesting that, unlike in polyolefin/wood-fiber composites, converting the hydrophilic
surface of wood fiber to hydrophobic is not effective for enhancing the adhesion of PVC

to wood fibers.

Previous studies, however, demonstrated that when PVC is used as the matrix in
WPCS, acid-base interactions, in which one phase acts as an electron donor (base) and the
other acts as an electron acceptor (acid), are a significant factor in enhancing interfacial
adhesion [2,3]. Therefore, surface modification of wood fibers to be used with PVC
should be designed to modify the acid-base interactions at the matrix/fiber interface in
order to improve the performance of these composites. For example, by changing the
acidic characteristics of wood fibers through surface modification with y-amino
propyltriethoxy silane (Figure 5.1), PVC/wood-fiber composite with equal tensile
strength and greater modulus than unfilled PVC was developed [2]. The use of the
aminosilane successfully modified the wood surface, and facilitated the interaction

between the wood and PVC according to Lewis acid-base theory [2].

In spite of these benefits, 'y-amino propyltriethoxy silane has not been extensively
used as a coupling agent for PVC/wood-fiber composites, mainly due to its high cost but
also due to the difficulty in evenly coating the surface of wood fibers, owing to its
sensitivity to hydrolysis and self-condensation. Consequently, aminosilane is not a

desirable coupling agent in this application.

94

NH2
. NH2 k
5: L /. 3’)
/I—OH /'——O-Si
+ Cl) —>
O-Sli
< “v

a : t: .
/' H /'—OH O\/
/| /'
/I
Wood Aminopropyltriethoxysilane Modified Wood
Cl
l\ NH2 \\ Nfi3— PVC
0 /i O >
é:__ Q_Sl j fl—O—Sli
V: '| + PVC —> f: (I)
| O /| \/
2 OH \/ /'—0H
/1 /l
Modified Wood

Figure 5.1. y-Amino Propyltriethoxy Silane Reaction Mechanism with Wood and
PVC [2]

95

Thus, the Objective of this study was to explore the use of amine-containing natural
polymers such as chitin and chitosan as cost-effective and novel coupling agents for

PVC/wood-fiber composites.

Literature Review on Chitin and Chitosan

Chitin (Figure 5.2a) is the second most abundant natural polymer after cellulose
and is extracted from the shells of crustaceans. Chitosan (Figure 5.2b) is obtained by N-
deacetylation of chitin. These polymers are widely available, non-toxic, biocompatible,
and lower in cost than many synthetic coupling agents. The acetyl amine functionality of
chitin, and the amine functionality of the chitosan, should permit these polymers to
interact with wood and PVC in a manner similar to the aminosilane, and so enhance the
interfacial adhesion between PVC and wood fibers, while potentially improving other

properties, and also be more cost-effective.

96

53339 Auv an: .5215 683:3 A3 he won—5935 .N.m 95$..—

£2

 

 

 

 

 

 

97

Several investigators have reported the use of chitin and chitosan as additives in
different polymers for various applications [10-16]. In one study, the interfacial adhesion
between chitin fibers and polycaprolactone (PCL) was increased by an irradiation
treatment of the composites. This resulted in an overall increase in mechanical properties
of the composite compared to composite prepared from untreated chitin fiber. This
increase in interfacial bonding was attributed to a free-radical grafting reaction [10].
Chitosan has also been reported to have been crosslinked to a polymer matrix [11].
However, that process used formaldehyde as a cross linking agent, which is a known

carcinogen and hazardous to the environment.

Biodegradable composites were also prepared by incorporating chitin flakes
ranging from 0-30 wt% into a plasticized starch matrix using an injection molding
process. Chitin flakes increased the elastic modulus, tensile stress and water resistance of
the composites when compared to the unfilled starch [12]. Biodegradation of the
synthetic polymers can also be increased by incorporating a natural biodegradable

polymer such as chitosan [13].

Chitosan is also used in the wood industry. For example, chitosan forms a Schiff
base when reacted with aldehyde compounds. This property of chitosan can be very
useful in reacting with the formaldehyde released from the glue line of plywood, thus
reducing the overall emission of formaldehyde to the environment. Hence, chitosan can
be used as a functional coating reagent for wood [15]. The use of chitosan as an

environmentally friendly adhesive for wood has also been reported in the literature. Glue

98

made from chitosan showed excellent water resistance and was proposed as a

replacement for synthetic adhesives [16].

Research has not been carried out to examine the effectiveness and efficiency of
chitin and chitosan as adhesion promoters in PVC/wood-fiber composites. These
compounds present an opportunity for compatibilizing PVC and wood-fiber in

composites in a way that would be novel, low cost and readily available.

99

EXPERIMENTAL

Materials

Chitin and chitosan, two natural polymers were used as coupling agents. PVC (K
value-66) was used with 40 mesh size wood flour from a hardwood maple species as
filler. The complete formulations of the rigid PVC/wood-flour composites are given in

Table 5.1.

100

Table 5.1. Formulations Used in PVC/Wood-Flour Composites.

 

 

Ingredients Content (phr)

PVC K-66 (OxyVinyls) 100
Tin stabilizer (PlastiStab 2808- from OMG Americas) 2

Calcium stearate (Synpro, Ferro Corp.) 1.5
Paraffin wax (Gulf Wax) 2
Paraloid K-120 (Rohm and Haas Co.) 2
Paraloid K-l75 (Rohm and Haas Co.) 2
Paraloid KM-334 (Rohm and Haas Co.) 10

Chitin or Chitosan (rcr America) Variable'
Wood flour (American Wood Fibers) 75

 

IThe concentrations of chitin and chitosan varied from 0 to 10 wt. % based on the weight

of wood flour in the composites.

101

Composite Manufacturing

All components Of the formulation given in Table 5.1 were added to a high
intensity mixer (Papenmeier, TGAHK20) and mixed at room temperature for 10 min.
Prior to compounding, the wood flour was dried in an oven for 48 hrs at 105°C to a
moisture content of less than 1%. The compounded and mixed formulation was then
extruded through a 32 mm conical counter rotating twin-screw extruder with an L/D ratio
of 13:1 (C. W. Brabender Instruments Inc.) into 10 mm diameter rods. The temperature
profile during extrusion was set at 190°C for all zones and the extrusion speed was
maintained at 40 rpm. The extruded rods were compression-molded into panels in a
hydraulic press (Erie Mill Co.) at 190°C for 5 minutes under 6200 KPa pressure. The

mold was then cooled to room temperature.

Property Testing

The samples for flexural testing were cut from the compression molded panels
and conditioned at 23 0C _+_ 2 and 50% i 5 relative humidity in a walk-in conditioning
room for 48 hours before testing. The flexural test was carried out on an Instron 4206
with series IX software as per ASTM D790. The crosshead speed was 1.9 mm/min. At
least eight samples were tested for each formulation. Tukey HSD test of multiple
comparisons was used to determine significant differences between the averages of the
batches. The effects of chitin and chitosan content on the flexural strength and flexural
modulus of the composites were compared at P < 0.05 to determine the statistical

significance.

102

Dynamic mechanical analysis (DMA) was carried out on a Perkin Elmer (DMA
7e) instrument in the three-point bending mode to determine the storage modulus (elastic
modulus), loss modulus (viscous modulus) and tan delta of the samples. The test was
performed in the temperature sweep mode from 25°C to 150°C under a helium

atmosphere at a frequency of 1 Hz and at a heating rate of 5°C/min.

103

RESULTS AND DISCUSSIONS

F lexural Properties

Figure 5.3 and 5.4 show the flexural strength and modulus of the neat rigid PVC

and composite without and with various coupling agent contents.

As expected, without coupling agent the addition of wood flour into the PVC
matrix reduced the flexural strength of the composite compared to the neat PVC (Fig.
5.3a and 5.3b). Poor interfacial adhesion between wood flour and PVC matrix may be

one of the reasons for the inferior flexural strength of the composites [4,6].

The incorporation of coupling agents into the formulations had a positive effect
on the flexural strength of PVC/wood-flour composites. Both chitin and chitosan
significantly improved the strength of the PVC/wood-flour composites compared to
composites without coupling agents (Figure 5.3a and 5.3b). Different letters above the
bars indicate significant differences between the uncoupled composite and composites
with chitin or chitosan, while the same letter indicates no statistically significant
difference at P < 0.05. The observed increase in flexural strength of the composites may
be attributed to the improved interfacial adhesion between the wood flour and the PVC
matrix [4,6]. This improved interface resulted in more effective stress transfer from the

matrix to the fiber.

The enhanced interfacial adhesion that occurs between wood flour and the PVC
matrix with the addition of coupling agents may be explained by the potential of chitin

and chitosan to alter the acid-base interactions at the interface between the wood flour

104

Flexural Strength (MPa)

 

 

 

 

 

 

 

 

 

 

60 ( )
a
B
B
55 - B
50 ~ 3111?; ii“
t:
45 - t—i‘s 213':
‘3': 1‘, 1 '1.
f
40 . ' T 1'12"; 1311-
PVC 0 2.5 5 6.6 7.5
Chitin content (%)
Figure 5.3.

 

 

 

Flexural Strength (MPa)

 

60

8 3
M 1

#
U1
1

 

 

 

 

 

.h
C

 

1.5 2.5
Chitosan content (%)

PVC 0 0.5 1

F lexural strength of PVC and its composites with varying
concentration of coupling agent: (a) chitin and (b) chitosan. Different
letters above the bars indicate significant differences between the
uncoupled composite and composites with chitin or chitosan at P <
0.05.

105

Flexural Modulus (GPa)

 

 

A
A

 

 

3.5- T I 33.5-
A O
1 In
2
31 g 3—
O
1 E
2.51 £2.5—
X
2
IL

  

4

 

 

 

 

 

 

 

 

 

 

 

13

   

 

N

‘17 1 1 7

PVC 0 2.5 5 6.67 7.5 10 PVC 0 0.5 1 1.5 2.5
Chitin content (%) Chitosan content (%)

Figure 5.4. Flexural modulus of PVC and its composites with varying
concentration of coupling agent: (a) chitin and (b) chitosan. Different
letters above the bars indicate significant differences between the
uncoupled composite and composites with chitin or chitosan at P <
0.05.

106

 

and the PVC matrix [2,3]. When these biopolymer coupling agents were incorporated
into the formulation, hydrogen bonds may have been formed between the hydroxyl
groups of the wood flour and chitin/chitosan coupling agents. As a result, the acidic
nature of the wood is moderated, and the wood surfaces that are bonded to the amino-
containing biopolymer have a more basic character. 0n processing, strong interfacial
adhesion between the chlorine-containing PVC and amino groups on the surface of wood
flour is promoted, which results in the significant increases measured in the mechanical
properties of the composites [2]. Chemically, chitosan also has the potential to form
covalent bonds with PVC through an 82.11 or SNZ reaction between the amine group of
chitosan and the chlorine atom of PVC, eliminating HCl. The covalent bond would
provide additional stability to the interface of the composite. This would likely not be a
dominant phenomenon since chitosan and PVC are both polymers, which increases both

the steric hindrance and the activation energy needed for the reaction.

The flexural strength of coupled composites increased with the addition of either
chitin or chitosan, with the extent of the increase depending on the amount of added
coupling agent. 0n addition of 6.67 wt. % of chitin to the composite the flexural strength
was approximately 54 MPa, compared to 45 MPa for the composite without a coupling
agent, and 57 MPa for the neat PVC (Figure 5.3a). Chitosan was more efficient in
enhancing the strength of PVC/wood-flour composites, giving a flexural strength of 53
MPa at 0.5 wt. % concentration, with no additional increase in properties attained on
addition of more chitosan (Figure 5.3b). This difference can be attributed to molecular
structures of chitosan that contain more reactive amino groups (primary amine) as

compared to chitin, which has less reactive acetylamine groups (Figure 5.2). In addition,

107

chitosan may have formed some direct covalent bonds to PVC, which would enhance the
flexural strength of the composite. Chitin cannot react in this manner due to the presence

of the less reactive secondary amine in the acetylamino group.

The flexural modulus was also significantly improved with the composite
containing 6.67 wt. % chitin having a flexural modulus of 3.7 GPa compared to 3.2 GPa
and 2.2 GPa for the uncoupled composite and neat rigid PVC, respectively (Figure 5.4a).
Again, the different letters above the bars indicate significant differences between the
uncoupled composite and composites with chitin or chitosan, while the same letter
indicates no statistically significant differences with the coupling agent at P < 0.05. The
composite modified with 0.5 wt. % chitosan had a flexural modulus of 3.5 GPa (Figure
5.4b). Without a coupling agent, the addition of wood flour into the PVC matrix
significantly increased the flexural modulus of the composites, as expected from the rule
of mixtures [5], but the addition of both chitin and chitosan significantly increased the
stiffness of the composites when compared to composites without coupling agent. This
enhanced stiffness is also attributed to the improved interfacial adhesion between the
PVC matrix and the wood flour. Other investigators have reported similar results [4].

Greater improvement in stiffness was observed with chitin in the composites.

Overall, the bending strength of rigid PVC/wood-flour composites with 6.67 wt.
% (~5.0 phr) chitin or 0.5 wt. % (~0.4 phr) chitosan matched that of unfilled PVC

whereas their stiffness outperformed that of neat PVC.
Dynamic Mechanical Properties

Figure 5.5 illustrates the DMA curve of composite without coupling agent.

Although not shown, similar curves were Obtained for neat rigid PVC and composites

108

with 6.67 wt. % chitin and 0.5 wt. % chitosan. These specific concentrations of coupling
agents for DMA study were selected on the basis of the optimum flexural properties
discussed in previous section. Table 5.2 summarizes the storage modulus (E’), loss
modulus (E”) and tan 5 peak max values for the neat rigid PVC, composite without and

with 6.67 wt% chitin and 0.5 wt% chitosan coupling agents obtained from DMA curves.

Both the storage modulus (E’ or elastic component of the material) and the loss
modulus (E” or viscous component of the material) increased with the addition of wood
flour into PVC matrix, regardless of both the type Of coupling agent used and testing
temperature. Increased storage modulus (E’) due to the addition of wood flour into PVC
matrix implies an increase in the load bearing capacity of the composites. This trend
correlates well with the flexural test data. The addition of wood flour increased the
viscosity of the polymer matrix, i.e. the loss modulus E”. It is believed that the presence

Of 75 phr wood flour in the PVC matrix restricted molecular motion of the matrix [5].

Composites prepared with chitin and chitosan showed a greater increase in both
storage modulus and loss modulus compared to the uncoupled composite. The increased
storage modulus (E’) is attributed to a better wood flour-matrix adhesion. This enhanced
interfacial adhesion may also restrict the free motion of the PVC chains, resulting in an
increased viscosity of the composites and the corresponding increase in loss modulus
(E”). Greater improvement in E’ and E” was observed with chitin in the composites. The
glass transition temperature (tan 5 peak max) of PVC was not significantly affected when
wood flour was added into the PVC matrix, irrespective of the type of coupling agent

used.

109

5.00 0.9

 

 

 

 

 

 

 

 

 

 

Storage Modulus
4.50 ~ 0.8
— - — - - Tan Delta

4.00 1 _ 0 7
’6 3.50
LL .
9 0.6
u, 3.00 1
g ~ 0.5
8 2.50 ~ '~._
5 ‘3 r 0.4
g 2.00 2 \\

s " 0.3

8 1.50 - x.
(I) ~ - _ . ,2 ______

1.00 - ~ 0.2

0.50 - - 0.1

0.00 . , , r 0

30 50 70 90 110 130

Tom pe rature (°C)

Figure 5.5. DMA curve of rigid PVC/wood-flour composite without coupling
agent.

110

Tan delta

Table 5.2. Effect of Coupling Agent on the Dynamic Mechanical Properties of Neat
Rigid PVC and PVC/Wood-Flour Composites.

 

Storage Modulus, E’ Loss Modulus, E” Tan 8 Peak

 

 

Samples (GPa) (GPa) Max

30°C 50°C 70°C 30°C 50°C 70°C (°C)

PVC 2.8 2.5 1.9 0.5 0.4 0.4 90.5

Composite 4.6 3.5 2.6 0.6 0.5 0.5 88.5
Composite with

6.67 wt. % chitin 8.0 5.9 3.8 1.3 0.9 0.8 90.9

compos‘te w‘th 0'5 6.1 5.2 3.5 1.0 0.9 0.8 90.5

wt. % chitosan

 

lll

CONCLUSIONS

Chitin and chitosan were used as coupling agents for rigid PVC/wood-flour
composites. Both polymers gave composites having a flexural strength greater than the
uncoupled composite and rivaling those of neat PVC, and had a flexural modulus that
exceeded the flexural modulus of both the PVC and the uncoupled PVC/wood—flour
composite. The chitosan was more effective in improving the properties compared to the
chitin, which may be due to the presence of more reactive amine groups in the backbone
of chitosan. The optimum wt. % for addition of chitin was 6.67 wt. %. Addition of
chitosan also improved the flexural strength and modulus of the composites up to 0.5 wt.
%, with higher amounts giving no additional increase. A DMA study showed that the
addition of chitin and chitosan also increased the E’ and E” of the composites. The
increase in properties by addition of chitin and chitosan was attributed to increased
interfacial adhesion between the modified wood surface and the PVC matrix, which may
occur through acid-base interactions between the chlorine-containing PVC and amino

groups on the surface of wood flour.

112

10.

REFERENCES

Clemons, C.M., "Wood Plastic Composite in the United States: The Interfacing of
Two Industries", Forest Products Journal, 52(6): 10-18 (2002).

Matuana, L.M., Woodhams, R.T., Balatinecz, J.J., and Park, C.B., "Influence of
interfacial interactions on the properties of PVC/cellulosic fiber composites",
Polymer Composites, 19(4): 446-455 (1998).

Matuana, L.M., Balatinecz, J .J., and Park, C.B., "Effect of surface properties on
the adhesion between PVC and wood veneer laminates", Polymer Engineering
and Science, 38(5): 765-773 (1998).

Li, Q. and Matuana, L.M., "Surface of cellulosic materials modified with
functionalized polyethylene coupling agents", Journal of Applied Polymer
Science, 88(2): 278-286 (2003).

Matuana, L.M., Park, C.B., and Balatinecz, J.J., "The effect of low levels of
plasticizer on the rheological and mechanical properties of polyvinyl
chloride/newsprint-fiber composites", Journal of Vinyl & Additive Technology,

3(4): 265-273 (1997).

Woodhams, R.T., Thomas, G., and Rodgers, D.K., "Wood fibers as reinforcing
fillers for polyolefins", Polymer Engineering & Science, 24(15): 1166-1171
(1984)

Kokta, B.V., Maldas, D., Daneault, C., and Beland, P., "Composites of poly(vinyl

chloride) - wood fibers. 111: Effect of silane as coupling agent", Journal of Vinyl
Technology, 12(3): 146—53 ( 1990).

Kokta, B.V., Maldas, D., Daneault, C., and Beland, P., "Composites of poly(vinyl
chloride)-wood fibers. 1. Effect of isocyanate as a bonding agent", Polymer-
Plastics Technology and Engineering, 29(1-2): 87-118 (1990).

Kokta, B.V., Maldas, D., Daneault, C., and Beland, P., "Composites of poly(vinyl
chloride) and wood fibers. Part II: Effect of chemical treatment", Polymer
Composites, 11(2): 84-9 (1990).

Yang, A. and Wu, R., "Enhancement of the mechanical properties and interfacial
interaction of a novel chitin-fiber—reinforced poly(e-caprolactone) composite by
irradiation treatment", Journal of Applied Polymer Science, 84(3): 486-492
(2002)

113

11.

12.

13.

14.

15.

16.

Zhang, Q., Liu, L., Ren, L., and Wang, F., "Preparation and characterization of
collagen-chitosan composites", Journal of Applied Polymer Science, 64(11):
2127-2130 (1997).

Souza Rosa, R.C.R. and Andrade, C.T., "Effect of chitin addition on injection-
molded thermoplastic corn starch", Journal of Applied Polymer Science, 92(4):
2706-2713 (2004).

Ratajska, M. and Boryniec, S., "Biodegradation of some natural polymers in
blends with polyolefines", Polymers for Advanced Technologies, 10(10): 625-633
(1999)

Thwe, M.M. and Liao, K., "Tensile behaviour of modified bamboo-glass fibre
reinforced hybrid composites", Plastics, Rubber and Composites, 31(10): 422-431
(2002)

Sato, K., Ota, H., and Omura, Y., "Development of functional coating reagent for
wood based materials by using chitosan", Advances in Chitin Science, 2: 897-901
(1997)

Umemura, K., Inoue, A., and Kawai, S., "Development of new natural polymer-
based wood adhesives 1: dry bond strength and water resistance of konjac
glucomannan, chitosan, and their composites", Journal of Wood Science, 49(3):
221-226 (2003).

114

CHAPTER 6

EFFECTS OF WOOD FLOUR AND CHITOSAN ON MECHANICAL,
CHEMICAL AND THERMAL PROPERTIES OF POLYLACTIDE

*This chapter is in press for publication in Polymer Composites (February 2007)

115

ABSTRACT

Composites of polylactide (PLA, 100-60 wt. %) and wood flour (0-40 wt. %)
were prepared to assess the effects of wood filler content on the mechanical, chemical,
thermal, and morphological properties of the composites. The polysaccharide chitosan (O
— 10 wt. %) was added as a potential coupling agent for the PLA-wood flour composites.
Addition of wood flour significantly increased the flexural modulus and the storage
modulus of PLA-wood flour composite, but neither the wood flour nor chitosan had an
effect on the glass transition temperature (T8). Fourier transform infrared (FTIR) spectra
did not show any evidence of covalent bonding, but chitosan at the interface between
wood and PLA is thought to have formed hydrogen bonds to PLA carbonyl groups. SEM
images of fracture surfaces showed fiber breakage was far more common than fiber
pullout in the composites. No evidence of discrete chitosan domains was seen in SEM
micrographs. When added at up to 10 wt. % (based on wood flour mass), chitosan
showed no significant effect on the mechanical, chemical, or thermal properties of the

composites, with property changes depending on wood flour content only.

Keywords: Polylactide, Thermal properties, FTIR, Coupling agent and Biocomposites

116

INTRODUCTION

Recently, biopolymers have become popular due to both increased social and
economic pressure to conserve petroleum resources. Biopolymers offer environmental
benefits such as biodegradability, reduced greenhouse gas emissions, and renewability of
the base material [1, 2]. Polylactide (PLA) has properties that are comparable to many
commodity polymers (e. g. PP, PE, PVC, PS) such as high stiffness, clarity, gloss and UV
stability [3, 4], but is a biobased polymer, often made from corn. PLA is still costly
compared to PP, but does have a low cost compared to other readily available
biopolymers that have similar properties.

Biopolymers in combination with natural fibers represent a growing share of the
composites market for various applications. The major advantages of using natural fibers
as reinforcements or fillers are their low cost, low density, high specific strength and less
abrasiveness to processing equipment compared to conventional inorganic fillers [5-7].
Natural fiber-filled biopolymers are often referred to as “green” composites as they are
completely biodegradable and renewable.

The mechanical properties of composite materials depend on several factors such
as the interfacial interaction between the matrix and fiber; type, aspect ratio and
concentration of fiber, etc. Although cellulosic fibers have been shown to be good
reinforcing agents for polymer composites, poor dispersion and weak interfacial bonding
of the hydrophilic fibers with the hydrophobic polymer matrix leads to poor strength
properties, compared to neat polymer. Poor dispersion is inherent to cellulosic materials
due to hydrogen bonding between the fibers but only weak dispersion forces between the

filler and the polymer [7-9].

117

Several researchers [10-15] have studied PLA composites with different non-
wood cellulosic fibers. Compared with unfilled PLA, decreased strength properties were
found for PLA composites made from different natural fibers [10, 15, 16]. This was a
function of natural fiber content and was attributed to poor interfacial adhesion between
the filler and matrix. Oksman and coworkers [13] found that PLA/flax fiber composites
have superior mechanical properties compared to PP/flax fiber composites, and hence
have the potential to replace PP/natural fiber composites in many industrial applications.
They also reported the need for better interfacial adhesion between the polymer matrix
and natural fibers to improve overall mechanical properties of the composites. Similar
results were obtained by other researchers when bamboo [l4], cellulose [15] and recycled
newsprint fibers [10] were used.

To overcome the poor interfacial adhesion between filler and matrix, researchers
using bamboo fiber esterified the fiber with maleic anhydride and successfully used this
for PLA/bamboo fiber composites [14]. Addition of 5% esterified bamboo fiber resulted
in nearly 21% improvement in the tensile strength and 10% improvement in tensile
modulus, while addition of more compatibilizer had no firrther effect on properties.
Although the esterified bamboo fiber increased the properties of the PLA composite, this
approach had several drawbacks including the additional step of manufacturing reactive
fibers before processing of the composites, and the use of a solvent-based process to
produce the reactive fibers, which is less desirable economically and environmentally. In
a similar, effort another researcher used maleated PLA as a coupling agent. Reported
results suggest this would be effective in PLA/natural fiber composites [17]. However,

maleated PLA is not available commercially and would be expensive and time

118

consuming to produce as a part of a composite manufacturing process. A commercially
available natural product is desirable as a coupling agent for “green” composites.

Our previous work showed that natural polymers such as chitin and chitosan can
be successfully used as coupling agents for PVC/wood-flour composites [18]. Addition
of chitosan and chitin increased the flexural strength and flexural modulus of PVC/wood-
flour composites as a result of increased interfacial adhesion between the polymer matrix
and the wood flour. This increase was attributed to acid-base interaction between the
PVC and coupling agent.

Unlike PVC, PLA cannot undergo acid-base interaction with chitosan, but Honma
et al. [19] demonstrated that both OH and NH; groups from chitosan (Figure 6.1) could
hydrogen bond with the carbonyl group of poly(e-caprolactone) (PCL). Both PCL and
PLA are aliphatic polyesters with significant structural similarity (Figure 6.1).
Furthermore, it is theoretically possible that a reaction could occur between the amine
groups of chitosan and the ester groups of PLA, forming a covalent bond between the
chitosan and PLA (Figure 6.2) [20]. Similarity between the chitosan and cellulose
structure (Figure 6.1) indicates these compounds should be highly compatible and readily
hydrogen bond with each other. We hypothesized that the addition of chitosan would
increase the interfacial adhesion between PLA and wood flour, resulting in a stronger
composite product.

Chitosan has not previously been studied as a wood-PLA coupling agent.
Furthermore, although several different cellulosic fibers have been used to manufacture

PLA composites, there have been limited studies that used wood as a filler for PLA

119

 

 

o (b)

 

 

 

 

 

Figure 6.1. Structures of (a) Polylactide (b) Poly(e-caprolactone) (c) Cellulose and
(b) Chitosan.

120

ii ii
C R' C
Heat
RNH2 + R/ \O/ R/ \NHR + H0 R'
Amine Ester

Figure 6.2. Possible Chemical Reaction between Amine and Ester Group

121

composites [21]. Wood fiber is readily available in the US as a waste product of other
wood manufacturing processes and is often very inexpensive when compared with other
natural fibers [6]. Although wood is similar to other cellulosic fibers there are also
differences, most notably in the ratios of lignin and hemicelluloses to cellulose [5]. Thus,
its strength is known to be different from those of other natural fibers [5].

Therefore, the specific goals addressed in this paper are: 1) determine the effects
of various concentrations of wood flour on the properties of PLA/wood composites and
2) evaluate the efficacy of chitosan as a coupling agent in PLA/wood composites. These
goals were addressed through analysis of the mechanical, chemical and thermal
properties of PLA/wood composites prepared with varying wood content (0 — 40 wt. %).
PLA/wood flour composites with 40 wt. % wood were then prepared with chitosan (0 —
10 wt. % based on wood flour content) and similarly tested to determine whether chitosan

was effective in improving the composite properties.

122

EXPERIMENTATION

Materials

Polylactide (PLA) 4042—D containing 94% L-lactide (Nature Works LLC,
Minnetonka, MN) was used as a matrix. Maple wood flour (4020) of 425 um average
size (American Wood Fibers, Schofield, WI) was used as a filler. Chitosan (TCI
America, Tokyo, Japan) was used as a coupling agent. Chitosan flakes were pulverized
into a powder (< 250 pm) in a Kleco 4200 pulverizer (Garcia Manufacturing, CA) at

room temperature before compounding. All other ingredients were used as supplied.

PLA/Wood Flour Composite Manufacturing

Before compounding, wood flour was dried in an oven for 48 hrs at 105 °C to a
moisture content of less than 1%, and PLA was dried at 75 °C for 24 hrs. Formulations
used 0-40 wt% of dried wood flour, 100-60 wt. % dried PLA, and chitosan added at 0-10
wt. % based on wood flour mass. The exact formulations are given in Table 6.1.

The polymer, wood flour and chitosan were weighed at the desired addition level
and premixed in a glass beaker. The mix was then melt compounded in a micro-
compounder (DSM Research, Netherlands) at 190 °C for 4 minutes at a mixing speed of
100 rpm. The micro-compounder had a total volume of 15 cc and was equipped with co-
rotating intermeshing conical twin screws of L/D ratio 18 to ensure homogeneous melt
mixing of the material. The molten compounded material was then immediately
transferred to a plunger-based mini-injection molding machine through a heated cylinder
to mold the test samples. Temperature of the heated cylinder was maintained at 183 °C,

while that of the mold was controlled at 40 °C.

123

Table 6.1. Individual Batch Composition

 

 

Experiment PLA Wood Flour Chitosanl
No. (%) (%) (%)
1 100 O 0
2 80 20 0
3 7O 30 O
4 6O 40 0
5 6O 4O 1
6 6O 40 2.5
7 6O 4O 5
8 6O 4O 10

 

' Chitosan content is based on the weight of wood flour

124

Mechanical Property Testing

An Instron 4206 (Norwood, MA) Universal testing machine equipped with series
IX software was used to test flexural properties of the samples as per ASTM standard
D790 [22]. The injection molded samples were conditioned at 50% i 5 relative humidity
and 23 °C : 2 in a walk-in conditioning room for 48 hours before testing. Sample
dimensions were 60 mm x 12 mm x 3 mm. The average of at least seven samples was

reported for each formulation.

Statistical Analysis

Tukey’s test of multiple comparisons was used to determine significant
differences between the averages of the batches. The effects of wood flour content and
chitosan content on the flexural strength and flexural modulus of the composites were

compared. P < 0.05 was used to determine statistical significance.

Dynamic Mechanical Analysis (DMA)

DMA was carried out on a TA Instruments DMA Q800 (New Castle, DE)
analyzer using a dual cantilever fixture to determine the storage modulus (elastic
modulus), loss modulus (viscous modulus) and tan 8 of the samples. The test was
performed in the temperature sweep mode from 30 °C to 100 °C at a frequency of 1 Hz
and a heating rate of 2°C/min. The sample dimensions were 60 mm x 12 mm x 2 mm.

At least three samples for each batch were tested.

125

Surface Analysis

FTIR spectroscopy was performed on a Shimadzu IRPrestige-21 (Columbia, MD)
spectrometer in attenuated total reflectance (ATR) mode. Samples were pulverized in a
Kleco-4200 pulverizer (Garcia Manufacturing, CA) to a powder, which was placed in the

sample holder. Each spectrum consists of 50 scans collected at a resolution of 4 cm"l

1

over the range of 4000 — 650 cm’ . Shimadzu IRsolution software was used to analyze

the collected spectra.

Scanning Electron Microscopy (SEA/I)

Fracture surfaces of tested (flexural test) specimens were observed using a JEOL
J SM-6400 (Tokyo, Japan) scanning electron microscope equipped with a LaB6 gun. An
accelerating voltage of 12KV was used. The samples were gold coated in a sputter coater

before collecting the images.

126

RESULTS AND DISCUSSION

Mechanical Properties

The addition of wood flour to the PLA matrix appears to be slightly reducing the
fiexural strength of the PLA/Wood flour composites, compared to unfilled PLA (Figure
6.3a). Based on Tukey’s t-test (P > 0.05 in all cases) this decrease was not statistically
significant. Interestingly, decreased strength properties, compared with unfilled PLA, has
been reported by other authors using natural fibers including bamboo, newsprint and
cellulose [10, 14, 15]. Since there was no significant drop in strength properties when
wood was added to PLA, wood appears to be a better filler for PLA than other natural
fibers. An increase in the standard deviation of the samples was observed as the wood
flour content increased, which was likely a result of reduced mixing efficiency at higher
wood flour loading, and could likely be decreased by increasing mixing intensity.

The addition of wood flour to the PLA matrix significantly increased its flexural
modulus, irrespective of the wood flour content used (Tukey’s t-test, P < 0.05 in all
cases) (Figure 6.3b). Adding 20 wt. % wood flour increased the flexural modulus of the
PLA from 3180 MPa to 4249 MPa, which represented an improvement of ~34%. At 40
wt. %, the modulus increase was approximately 87% compared to unfilled PLA. This
behavior was expected, based on the rule of mixtures, as the modulus of wood fiber is
higher than that of the polymer.

The rule of mixtures is a mathematical equation that relates a property of a
composite to the properties of each component and volume fractions of the components
in the composite. Using the rule of mixtures, the modulus of the composite product can

be estimated by the following equation [23]:

127

 

 

EC = Ef Vf + Enr(1'vf)

where E, Em and Er are the moduli of composite, matrix and fiber respectively and Vf is
the volume fraction of fiber. The modulus of fiber used in the calculation was 20 GPa,
which represents a general value for wood fiber [23].

The flexural modulus of the composite increases with increasing fiber volume and
follows the predicted trend, however at much lower values (Figure 6.4). The prediction
from the rule of mixtures is based on several assumptions such as: fibers are uniform and
unidirectionaly oriented, perfect bonding between the fiber and matrix exists, and applied
load produces equal strain in the fiber and matrix. In the experimental samples, it is
unlikely that the assumptions hold true, which accounts for the decreased values
compared to those predicted for the system. Weak interfacial adhesion between the wood
flour and the matrix, uneven size of the wood flour used in this work, and random
orientation of the fibers in the composite were likely responsible for the experimental
values being lower than predicted. An increase in the standard deviation of the samples
was also observed for flexural modulus as the wood flour content increased (Figure 6.3b)
which was likely a result of reduced mixing efficiency at higher wood flour loading.

Chitosan had no significant effect on either the flexural strength (Figure 6.5a) or
on the modulus (Figure 6.5b) of the composite (Tukey’s t-test, P > 0.05 in all cases).
This implied that chitosan was not a suitable coupling agent for PLA-wood composite,
possibly due to incompatibility between chitosan and PLA. An incompatibility between
chitosan and PLA was reported by Suyatma et al. [24]. They studied blends of

chitosan and PLA prepared at room temperature by a solution blending technique. The

128

105 (a)

100

   

  
 

I‘
0‘ '

      

”urn: '
.‘i/fzz A.
3’;

V

   

('5?

l
..)r..

95

         
    
  

- v

  
    

.7')‘.". l .

NRA?! ‘v

-.I-.-v..-
1).. .

90

85

Flexural Strength (MPa)

 

 

 

 

 

 

80

7000

ON
0
O
O

5000

4000

 

3000

Flexural Modulus (MPa)

 

 

 

2000

 

Wood Flour Content (%)

Figure 6.3. Flexural Properties of PLA and its Composites with Varying
Concentration of Wood Flour (a) Flexural Strength and (b) Flexural
Modulus. Different letters above the bars indicate statistically significant
differences at P < 0.05.

129

 

   
  

 

 

 

 

 

 

 

 

10000
A 9000‘ APredictcd
a .
E 8000i IExpcnmcntal
3 7000-
3
e 6000~
2
3 5000‘
2
.2 4000‘
Ti.
3000
2000 . I I
0 10 20 30 40

Fiber Volume (%)

Figure 6.4. Predicted and Experimental Values of Flexural Modulus with Increasing
Volume of Fiber

130

 

 

 

 

85

Flexural Strength (MPa)

 

 

 

 

. 9 o
$.121'o:o:¢
4‘9 ’0 o o

O 1 2.5 5 10
Chitosan Content (%)

 

 

80r

 

7000

6000

5000

4000

Flexural Modulus (MPa)
Lg
G
O

 

2000

0 1 2.5 5 10
Chitosan Content (%)

Figure 6.5. Flexural Properties of PLA/Wood Flour Composites with 40 wt. %
Wood Flour and Varying Concentration of Chitosan (a) Flexural
Strength and (b) Flexural Modulus. Different letters above the bars
indicate statistically significant differences at P < 0.05.

131

content of PLA in these blends was varied from 0 — 30 wt. %. Under these conditions,
the chitosan/PLA blends were found to be incompatible. The temperature used to
manufacture these blends may not have been sufficient for a reaction to occur between
the chitosan and PLA. In the current work, an extrusion process at elevated temperature
and high pressure was used, which was expected to provide better conditions for a
reaction or interaction between PLA and chitosan. However, results of flexural testing

did not support this.

F TIR Analysis

Several characteristic peaks representing chemical functional groups were
detected in the FTIR spectra of PLA, wood and the wood/PLA composite with and
without chitosan (Figure 6.6, Table 6.2). The strong peak in the spectrum of PLA (Figure
6.6a) at 1747 cm'I is due to C=O stretching of the carbonyl group, while the bending
vibration of this group appears at 1265 cm". The peaks at 866 and 754 cm'l represent the
amorphous and crystalline phases of PLA, respectively. It is interesting to note that the
intensity of the amorphous and crystalline peaks is nearly the same, which indicates the
PLA is a semicrystalline material with nearly equal presence of both phases.

Compared to the spectrum of neat PLA, the spectrum of the composite containing
40 wt. % wood flour shows the presence of OH groups from the wood at 3332 cm'l
(Figure 6.6c). Some reduction in peak intensity of the characteristic PLA peaks was
expected in the composite spectrum as a result of wood flour addition, simply because of
a dilution effect. However, comparing the ratios of characteristic peak intensity of PLA

to the composite spectrum indicated that the intensity of the peak at 1749 cm'l in the

132

Table 6.2. FT IR Band Assignment of Major Peaks in PLA and Wood Flour

 

 

(1:113) 22:)"; Peak assignments References
--- 3331 OH stretching [33, 34]
2993, 2943 2895 CH stretching [3, 4, 33]
1747 1735 C=O stretching [4, 33, 34]
1450 --- CH3 bending [3, 4, 34]
1381, 1357 1400-1300 CH deformations [3, 4, 33, 34]
1265 --- C=O bending [3, 4, 34]
---- 1232 phenolic OH deformation [33, 34]
1 180, 1126, 1082 --- C-O stretching [3, 4, 34]
1041 --- OH bending [3, 4, 33, 34]
1028 s...ii.‘3.;i‘%fi'€;.filng [33,341
866 --- PLA amorphous phase [3, 4]
754 --- PLA crystalline phase [3, 4]

 

133

 

 

 

 

 

 

 

 

 

 

I 'U
- E
- a
" 7S1. : h
75:. —8 =
Pm 2
998 L98 t [a
t '8
I O Q
r§ B
”’0! 8101 [701 ON" : °\°
1801 280! I801 : .
9le E0 E r
081! 18” ~53 3 an
ICU :‘— .5 ’
99zr _ ._ .‘
: 3 : at
8991 :8 3 a ..
tact __v E 8 a
- ‘— «H .
"SN ZS“ : as 4.1,;
t §u
Io U \9
‘ 1(91 _0 o
[651 l _ 9 < g
[5 u—l
Z 9" c
‘ l— AF!
6%! ”83 8:. a
: fl e ‘5
: '5 Ln
lg 8 e
- o
:N 3 o
- a 3
L' %3
—§ 2 .3
IN A 3
I as
I < 3
. O _J a
_O n‘ '-
'8 A 3
”a 96’“ CV62 5 3g
- m
£662 £66: _ .5 o
:O a :-
E
:8 3: g
- m 3 U
[995 ZCCC - D.
E m <
. o M 3]
e8 i: A
y.
:m In 3
" ' A . a E If;
3’ E 8.- ‘V :81 2
ll1lllTllllllllll]lllldlrlllllIlllllllIlll'ITllllllllllllllllllIllllflIlll 9 O?”
aoueqsosqv Ln

 

134

composite was further reduced, beyond that expected by dilution. In addition, the
carbonyl peak was shifted very slightly compared to PLA alone (1747 to 1749 cm").
This may indicate an interaction between the carbonyl groups of PLA and hydroxyl
groups of wood through hydrogen bonding [25-27]. The decreased intensity of the
crystalline peak of PLA at 754 cm"1 with respect to the amorphous peak at 867 cm'l
suggests a reduction in the crystallinity of PLA with the addition of wood flour. Reduced
crystallinity of PLA by addition of wood or other natural fibers has been reported in the
literature [10, 21]. The composite spectrum did not show any new peaks that were not in
the original wood or PLA spectra. Similarly, no major shifting or disappearance of peaks
was observed. The composite spectrum was essentially an overlay image of both PLA
and wood flour spectra, which was expected because the two dissimilar phases in the
composite were only likely to interact through hydrogen bonding.

Similarly, addition of chitosan had no effect on the FTIR spectra of the composite
(Figure 6.6d). The IR spectrum of the composite with 10 wt. % chitosan looked identical
to the spectrum of the composite with no chitosan (i.e. Figure 6.6c). There was no
evidence of chemical bonding between the amine group of chitosan and the ester
carbonyl group of PLA (Figure 6.2), which should have been detected as a peak near
1650 cm'l for the amide carbonyl. This finding is in agreement with the flexural
properties of the composites, where chitosan had no significant effect on mechanical
properties. However, we were also not able to detect any of the primary characteristics of
chitosan, such as the presence of amino groups (Figure 6.1), which would appear between
1600-1500 cm'I in the spectrum of the composite with chitosan. This might be due to the

low concentration of chitosan used in the study, or due to limitations of the ATR

135

technique, which is highly surface sensitive, as the IR beam can only penetrate the

sample 0.5-3 urn depending upon the refractive index of the sample and crystal [28, 29].

Dynamic Mechanical Analysis

The temperature dependence of storage modulus and tan delta for unfilled PLA
(curve a) and PLA wood composite with 20 wt. % (curve b) and 40 wt. % wood flour
(curve c) is illustrated in Figure 67(1). The storage modulus significantly increased as
wood flour was added to the PLA. This trend was similar to the results obtained in the
fiexural test and was expected on the basis of the rule of mixtures. Above T8, at around
80 °C, the storage modulus starts to slightly increase again in both unfilled and filled
PLA, due to a phenomenon known as cold crystallization [13]. This behavior in unfilled
PLA cannot be seen because of the scale used (Figure 6.7(I)). During molding, the
samples were quickly cooled from 183 °C (processing temperature) to 40 °C (mold
temperature), which restricted the chain movement and thus prevented the formation of
crystallites. When these samples were again heated above Tg in the DMA test, the
molecular motion within the chain increased and they tended to rearrange themselves to
form crystallites. This process is referred to as cold crystallization. This increase in
crystallinity gives added strength to the material, and thus an increase in the storage
modulus.

The loss modulus (Figure 6.7(II)) peak represents the glass transition temperature
(Tg) of the polymer. Addition of wood flour changed the Tg of the polymer by less than 1
°C. A similar trend of increased storage modulus with no significant change in the Tg of

the PLA was observed by Huda et al. when cellulose fibers were added to PLA [13]. As

136

7000 2.0

(1) .

 

\'\ / -1.5

2
Tan Delta

Storage Modulus (MPa)
E

E
I
~>*~. if ;

.A.
/,/
r
c:
on

 

1000- . \

 

 

 

V - 1 j T V I V I 1 Th— ' _h‘ V I 0.0

0
20 30 40 50 60 70 80 90

 

1200
10005 [ l / (C)
800:

600 ~

Loss Modulus (MPa)

400*

200 .

 

 

 

 

Temperature (C)

Figure 6.7. Temperature Dependence of (1) Storage Modulus and Tan Delta and (11)
Loss Modulus of (a) PLA, (b) PLA Composite with 20 wt. % Wood
Flour and (c) PLA Composite with 40 wt. °/o Wood Flour.

137

 

 

 

the Tg occurred over a range of temperature, it can be seen from the graphs that addition
of wood flour broadened the loss modulus curves and increased the glass transition
period. This was likely the result of disruption of the movement of the polymer chains by
added wood flour. Presence of rigid wood flour hinders the large-scale motion of the
polymer chains, which may lengthens the time of the transition from glassy to rubbery
state [30]. Similar to the storage modulus, the loss modulus increased when the
increasing content of wood in the composite. The..increasing trendflin 1055-59981335 ‘
means that the composite can dissipate more energy into heat without deforming, which -
would indicate that the composite has greater resistance to flow under conditions of stress 1
compared with unfilled PLA at Tg [31].

The tan 8 peak height decreased significantly with the addition of wood flour, and
showed a continued decreasing trend as the wood content was increased from 20 to 40
wt. % in the composite (Figure 6.7(I)). Tan 8 was reduced as the addition of wood flour
reduced the free volume of the polymer, restricting the molecular motion of the polymer
chains. “However, there was no change in the tan 8 max temperature. This signifies that:
there was no strong interaction between wood and PLA, which correlates with the
findings of the fiexural and FTIR analysis. Additionally, the width of the tan 8 peak
increased with increased percentage of the fiber in the composite. Idicula et a1. attributed
this phenomena to molecular relaxations taking place in the composite which were not
present in the matrix [32].

There was no significant change in storage modulus, tan delta or loss modulus
with addition of chitosan to PLA/wood composite, irrespective of the amount added

(Figure 6.8). This indicates that chitosan was not an effective coupling agent for PLA-

138

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

7000 1.0
(1) ~
4 (b) [A' ..
(b) I i
6000i 4 --------- if (a) .
/ \ ‘7' ‘ - 0.8
r \
4: ‘.<——(c)
A 5000- (C) a
(Q (a) l l \
8 ~ ‘ r = ~
I ‘ ’i II \ " 0.6
3 4000- 1 Ir . _ a
5 l . _
3 it i 8 Ir-
2 i] .’ L 5
o 3000‘ ' I—
m — 0.4
a: l
5 i l l '
a: 2000- i. l, r ,
1' l . .
t.’ r .
l l '- 02 a
’l \\ "”4
1000- /, \\ J
l \ "‘3’
I \ / .
0 I r l I T I ' l I 0.0
20 30 40 60 70 80 90 100
Temperature (‘Cl
‘ II
1200* ( )
1000~
E .
5 800—
a .
2 -l
3 4
o 600~
E .
_r .
400-
200-
ll _ f
o I I l T T j I ' I '
20 30 40 60 70 80 90 100

Temperature (C)

Figure 6.8. Temperature Dependence of (1) Storage Modulus and Tan Delta and

(11) Loss Modulus of PLA composite with 40 wt. % Wood Flour and
Varying Amount of Chitosan (a) 0%, (b) 2.5% and (c) 5%.

139

wood flour composites. This result agrees with the results of the flexural testing and
FTIR analysis which indicated that there was no significant effect of adding the chitosan

to the composites.

Morphology

The SEM study was performed on fracture surfaces of samples from the flexural
test. SEM images of PLA/wood flour composites containing 20 wt. % wood flour
(Figure 6.9) were recorded at various locations across the fracture surface.
Representative images from two different areas of the fracture surface are shown (Figure
6.9a and 6.9c). The SEM images suggest that the fibers were well dispersed in the matrix
resulting in a homogenous composite. Several fibers can be seen in these images, many
of which are broken at the surface of the sample. To better understand the mode of
failure in the fiexural samples, an area of each image (rectangle in Figures 6.9a and 6.9c)
was firrther magnified to display the wood particles in greater detail (Figures 6% and
6.9d). Higher magnification (Figure 6%) shows two wood particles which have been
broken off evenly with the surface of the polymer matrix. Fiber breakage indicates good
interfacial adhesion between the wood and PLA matrix. Conversely, pullout of fibers
from the matrix would indicate poor adhesion between the wood and polymer. There are
some areas (Figures 6.9a and 6.90) that show fiber pullout, but those are a minority
compared with the number of fibers that were broken. Even the fibers which were pulled

out left distinct imprints of the wood fiber surface instead of clean pullout in the matrix,

140

 

 

'-. (a)

' .45.
" Fiber bfealmge-——-———*

 

Figure 6.9. SEM Images of PLA Composites with 20 wt. % Wood Flour at Different
Magnifications: (a, c) 100X and (b, d) 400x.

141

suggesting mechanical interlocking between the fiber and matrix. Good mechanical
interlocking between PLA and wood flour was also observed through SEM recently by
Mathew et a1. [21] and was attributed to the surface roughness of the wood flour. An
additional feature (Figure 6.9d) is what appears to be polymer filling the lumens of the
wood fiber, which can be seen in the broken fiber. The polymer appears as parallel lines
that seem to be holding the halves of the broken fiber together. If the PLA is able to flow
into the hollow lumens of the wood particles and fill them, this would add strength to the
composite. This, along with mechanical interlocking of the filler into the matrix, is likely
responsible for the insignificant drop in flexural properties when wood was added to the
PLA matrix.

The fracture surfaces of PLA/wood flour composites with 40 wt. % wood (Fig.
10) without chitosan (Figures 6.10 a and b) and with chitosan (Figures 6.10 c and d) show
similar characteristics to those with 20 wt. % wood (Figure 6.9), i.e. more fiber breakage
was seen than fiber pullout. Chitosan could not be differentiated from small particles of
wood flour in these images, which may be a result of the small size of the chitosan
particles used (less than 250 microns). Images of the fracture surface at higher
magnification (Figure 6.10b and 6.10d) compared to the same areas under low
magnification (Figure 6.10a and 6.10c) more clearly show fiber breakage both across the
length and parallel to the length of the fibers. These SEM observations further support
the findings from the flexural study that addition of wood had no destructive effect on the
strength properties of the PLA, indicating that wood is a suitable filler for PLA in

composites.

142

 

 

Figure 6.10. SEM Images of PLA Composite with 40 wt. % Wood Flour without (a,
b) and with 10 wt. % Chitosan (c, d) at Different Magnifications: (a, c)
100X and a), d) 400x.

143

 

 

CONCLUSIONS

Results of this study indicate that wood flour can be successfully used as a filler
for PLA. The resulting composites will be renewable and biodegradable. The storage
and flexural modulus of the PLA/wood composites were higher than those of unfilled
PLA, regardless of wood content, but the strength properties were not improved by
adding wood. FTIR analysis of the composites produced spectra that were essentially an
overlay of the two component spectra, suggesting little interaction between the phases.
The SEM study provided evidence that there was sufficient interfacial adhesion between
the PLA matrix and the wood filler to cause the composite to fail mostly through fiber
breakage, rather than fiber pullout. This was likely a result of mechanical interlocking
between the wood fibers and PLA matrix, and the observed filling of the wood lumens by
PLA.

Chitosan was tested as a natural, biodegradable coupling agent for PLA/wood
flour composites. The results of this study did not show any significant improvement in
the mechanical or thermal properties of the composites when chitosan was added. No
differences were detected in the FTIR spectra of composites with chitosan, compared
with composite spectra without chitosan. The lack of differences in the spectra suggest
that no chemical bonding occurred between the wood and PLA. Since there were no
improvements when chitosan was used, we can conclude that chitosan was not a suitable
coupling agent for PLA/wood flour composites. However, PLA/wood flour composites

without coupling agent have enhanced moduli compared to PLA alone.

144

 

 

10.

11.

12.

REFERENCES

. Chandra, R. and Rustgi, R., "Biodegradable polymers", Progress in Polymer Science,

23(7): 1273—1335 (1998).

Van de Velde, K. and Kiekens, P., "Biopolymers: overview of several properties and
consequences on their applications", Polymer Testing, 21(4): 433-442 (2002)..

Garlotta, D., "A Literature Review of Poly(Lactic Acid)", Journal of Polymers and
the Environment, V9(2): 63-84 (2001).

Auras, R., Harte, B., and Selke, S., "An Overview of Polylactides as Packaging
Materials", Macromolecular Bioscience, 4(9): 835-864 (2004).

Bledzki, AK. and Gassan, J., "Composites reinforced with cellulose based fibres",
Progress in Polymer Science, 24(2): 221-274 (1999).

Clemons, C.M., "Wood Plastic Composite in the United States: The Interfacing of
Two Industries", Forest Products Journal, 52(6): 10-18 (2002).

Saheb, D.N. and Jog, J .P., "Natural fiber polymer composites: A review", Advances
in Polymer Technology, 18(4): 351-363 (1999).

Kazayawoko, M., Balatinecz, J.J., and Matuana, L.M., "Surface modification and
adhesion mechanisms in wood fiber-polypropylene composites", Journal of Materials
Science, 34(24): 6189-6199 (1999).

Felix, J.M. and Gatenholm, P., "The nature of adhesion in composites of modified
cellulose fibers and polypropylene", Journal of Applied Polymer Science, 42(3): 609-
620 (1991).

Huda, M.S., Drzal, L.T., Misra, M., Mohanty, A.K., Williams, K., and Mielewski,
D.F., "A Study on Biocomposites from Recycled Newspaper Fiber and Poly(lactic
acid)", Industrial & Engineering Chemistry Research, 44(15): 5593-5601 (2005).

Hou, Q.X., Chai, X.S., Yang, R., Elder, T., and Ragauskas, A.J., "Characterization of
lignocellulosic-poly(lactic acid) reinforced composites", Journal of Applied Polymer
Science, 99(4): 1346-1349 (2006).

Shanks, R.A., Hodzic, A., and Ridderhof, D., "Composites of poly(lactic acid) with
flax fibers modified by interstitial polymerization", Journal of Applied Polymer
Science, 99(5): 2305-2313 (2006).

145

 

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

Oksman, K., Skrifvars, M., and Selin, J.F., "Natural fibres as reinforcement in
polylactic acid (PLA) composites", Composites Science and Technology, 63(9): 1317-
1324 (2003).

Lee, S. -.,H Ohkita, T., and Kitagawa, K. "Eco- c-omposite from poly(lactic acid) and
bamboo fiber", Holzforschung, 58(5): 529-536 (2004).

Huda, M.S., Mohanty, A.K., Drzal, L.T., Schut, E., and Misra, M., ""Green"
composites from recycled cellulose and poly(lactic acid): Physico-mechanical and
morphological properties evaluation", Journal of Materials Science, 40(16): 4221-

4229(2005)

Lee, S.-H. and Wang, S., "Biodegradable polymers/bamboo fiber biocomposite with
bio-based coupling agent", Composites, Part A: Applied Science and Manufacturing,
37(1): 80-91 (2006).

Plackett, D., "Maleated Polylactide as an Interfacial Compatibilizer in
Biocomposites", Journal of Polymers and the Environment, 12(3): 131-138 (2004).

Shah, B.L., Matuana, L.M., and Heiden, P.A., "Novel coupling agents for PVC/wood-
flour composites", Journal of Vinyl & Additive Technology, 11(4): 160-165 (2005).

Honma, T., Senda, T., and Inoue, Y., "Thermal properties and crystallization
behaviour of blends of poly(e-caprolactone) with chitin and chitosan", Polymer
International, 52(12): 1839-1846 (2003).

Popoola, V.A., "Polyester formation: Aminolytic degradation and proposed
mechanisms of the reaction", Journal of Applied Polymer Science, 36(7): 1677-1683
(1988)

Mathew, A.P., Oksman, K., and Sain, M., "Mechanical properties of biodegradable
composites from poly lactic acid (PLA) and microcrystalline cellulose (MCC)",
Journal of Applied Polymer Science, 97(5): 2014-2025 (2005)._

ASTM-D790, Standard T est Methods for F lexural Properties of Unreinforced and
Reinforced Plastics and Electrical Insulating Materials, West Conshohocken, PA:
ASTM International (2003).

Simonsen, J ., "Efficiency of reinforcing materials in filled polymer composites",
Forest Products Journal, 47(1): 74-81 (1997).

146

 

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

Suyatma, N.E., Copinet, A., Tighzert, L., and Coma, V., "Mechanical and Barrier
Properties of Biodegradable Films Made from Chitosan and Poly (Lactic Acid)
Blends", Journal of Polymers and the Environment, 12(1): 1-6 (2003).

Yoshie, N., Azuma, Y., Sakurai, M., and Inoue, Y., "Crystallization and compatibility
of poly(vinyl alcohol)/poly(3-hydroxybutyrate) blends: Influence of blend
composition and tacticity of poly(vinyl alcohol)", Journal of Applied Polymer
Science, 56(1): 17-24 (1995).

Zhang, G., Zhang, J ., Zhou, X., and Shen, D., "Miscibility and phase structure of
binary blends of polylactide and poly(vinylpyrrolidone)", Journal of Applied Polymer
Science, 88(4): 973-979 (2003).

Shuai, X., He, Y., Asakawa, N., and Inoue, Y., "Miscibility and Phase Structure of
Binary Blends of Poly(L-lactide) and Poly(vinyl alcohol)", Journal of Applied
Polymer Science, 81(3): 762-772 (2001).

Silverstein, RM. and Webster, F .X., Spectrometric Identification of Organic
Compounds. 6th ed, New York: John Wiley. xiv, 482 p. (1998).

Vickerman, J .C., Surface analysis .° The Principal Techniques, Chichester, New York:
John Wiley. xvi, 457 p. (1997).

Tajvidi, M., "Static and dynamic mechanical properties of a kenaf fiber-wood
flour/polypropylene hybrid composite", Journal of Applied Polymer Science, 98(2):
665-672 (2005).

Menard, K.P., Dynamic mechanical analysis: a practical introduction, CRC Press:
Boca Raton, FL. 208 p. (1999).

Idicula, M., Malhotra, S.K., Joseph, K., and Thomas, S., "Dynamic mechanical
analysis of randomly oriented intimately mixed short banana/sisal hybrid fibre
reinforced polyester", Composites Science and Technology, 65(7): 1077-1087 (2005).

Marcovich, N.E., Reboredo, M.M., and Aranguren, M.I., "FTIR spectroscopy applied
to wood flour", Composite Interfaces, 4(3): 119-132 (1996).

Socrates, 6., Infrared Characteristic Group Frequencies: Tables and Charts. 2nd ed,
Chichester, New York: John Wiley. viii, 249 p. (1994).

147

 

 

CHAPTER 7

SUMMARY AND RECOMMENDATIONS

Summary

The aim of this work was to advance wood composites technology and utilization
through experimentation focused on improving WPCS properties and broadening the
applications for WPCS. Research was focused on three goals which address current
limitations of WPCS; i) development of a continuous extrusion process for microcellular
foaming of polyvinyl chloride (PVC) and its composites using supercritical fluids to
reduce the high density of the WPCS, ii) development of coupling agents for PVC and
polylactide (PLA) based WPCS to overcome the poor compatibility between the wood
and plastic, and iii) demonstrate that wood can be successfiilly used as a filler for
biodegradable WPCS based on PLA. Reaching these goals will open up new areas for

WPC utilization and expand composites technology.

The following conclusions were drawn from this work:

Complex relationships between wood flour, acrylic modifier and plasticizer on the
die swell and true viscosity of PVC and PVC wood flour composites have been identified
and explored. Understanding the relationships between key processing variables is
critical in producing high-quality products, and is vital for implementing an extrusion
processing method for microcellular foaming.

A greater understanding of the conditions necessary for foaming PVC and its
wood composites was developed. Although PVC was not microcellular foamed in the

extrusion system using supercritical CD; as the blowing agent, understanding the role of

148

 

1““

 

mixing and pressure on the PVC foam has lead to the generation of new ideas on how to
successfully foam this polymer.

Chitin and chitosan were successfully used to compatibilize PVC wood flour
composites. These efficient, cost effective coupling agents improved the compatibility
between the wood and PVC, resulting in better composite properties. Although both
compounds improved the composite properties, chitosan was more effective than chitin.
Use of these coupling agents can improve PVC wood composite products at a lower cost
and they are simpler to use than the current alternatives.

This work demonstrated that wood was an effective filler for PLA composites.
Adding wood to the PLA increased both the flexural and storage moduli, compared to
unfilled PLA. These composites are natural and biodegradable, and will expand the use
of WPCS in applications such as packaging and the automotive industry. Chitosan did
not improve the properties of the composites, but this investigation lead to greater
information about PLA reactivity with amine containing polymers.

Taken together, these results have added to the current state of the knowledge on
both PVC and PLA wood flour composites. The property improvements and increased
understanding of wood polymer composites meet our original aim of advancing wood
composites technology. This knowledge increases the potential applications for WPCS in

new areas. Information gained from this work also provides a basis for future studies.

149

 

 

Recommendations

Based on the results of this research, the following recommendations for future work are

suggested:

1.

The viscosity measurements done as part of the rheology study were performed in
an extrusion capillary rheometer. This type of rheometer is known to operate at
high shear rates. A complimentary study at low shear rates should be performed
using a conventional rheometer, such as cone and plate, to fully understand the

effects of material composition over a wide range of shear rates.

Results of the rheological study revealed complex interactions between acrylic
foam modifier and components of the PVC wood flour composite. The behavior
in the composite was opposite from that observed in the unfilled PVC. In the
PVC, melt viscosity was observed to increase, which is typical behavior when
acrylic foam modifiers are added to PVC. The reasons for the decreased viscosity
in the composite with acrylic foam modifier are not well understood, and should

be investigated in detail.

The foaming setup developed as a part of this work was not effective for
producing microcellular foamed PVC at steady-state conditions, even though
microcellular foams of other polymers were successfully produced using this
equipment. This was likely due to insufficient mixing of polymer and blowing
gas to produce microcellular foam. A setup utilizing an extruder with a larger
L/D ratio could improve the foamability of PVC with supercritical C02 as the

blowing agent. Another suggestion would be to use a different flow restrictor

150

 

which would maintain a higher pressure differential between the barrel and the
blowing gas. Better control of the static mixer temperature may also improve the

ability to produce microcellular foamed PVC with this setup.

. When chitin and chitosan were used as coupling agents for PVC wood flour
composites, improved properties were observed. The increase in the properties
was attributed to possible acid-base interactions between the coupling agents and
the PVC wood flour composites. A detailed investigation of surface properties of
the composites with and without coupling agent should be carried out to better
characterize the mechanism for improved adhesion. Inverse phase gas
chromatography or contact angle measurements could be used for these

experiments.

. Chitin and chitosan are known as anti-fungal agents. This aspect of chitin and
chitosan was not within the scope of this research. A study could be done to
assess the effectiveness of chitin and chitosan as anti-fungal agents for PVC wood

composites.

. In the PLA-wood flour composite study, 40-mesh size maple flour was used as
the filler for PLA. This wood flour was an effective filler for PLA. The effects of
other mesh sizes and species of wood flour on the properties of PLA are unknown

and should be studied.

. One goal of the PLA-wood flour composite study was to evaluate chitosan as a

natural coupling agent for these composites. While the results of the study

151

 

,‘4

 

reaffirmed the need for an effective coupling agent for these composites, chitosan
did not improve the composite properties. Maleated polymers are known to be
effective couplers for other composite systems. Maleated PLA should be

investigated as a coupling agent to improve the properties of these composites.

152

 

APPENDIX A

Wood:

Wood is one of the major natural fibers used in making composites and is also a
focus of this research. Wood is used in several forms in manufacturing composites such
as flour (particles), flakes, fibers and pulp. Wood is a cellular, fibrous material available
abundantly in nature and can be broadly classified as softwood and hardwood [1, 2]. The
wood from gymnosperm (conifer) trees is generally called softwood, e.g. pine, cedar,
spruce, larch, douglas fir, etc. The wood from angiosperm (deciduous) trees is generally
referred to as hardwood, e.g. maple, oak, teak, aspen, ash, etc. The terms softwood and
hardwood should not be confused with the literal meaning of soft and hard, as some
softwoods are harder and denser than hardwood and some hardwoods are softer than
softwood [1, 3]. The structure and chemical composition of the wood varies greatly with
species. Table A.l shows the difference in chemical compositions of softwood and

hardwood.

Table A.1. Chemical Composition of Wood

 

 

 

 

 

Chemical Component Sofét;:r)1od Haze/300d
Cellulose 40-45 45—50
Hemicellulose
i. galactoglucomannan 15-20 -

ii. arabinoglucuronoxylan 10 -

iii. glucuronoxylan - 20-30
iv. glucomannan - 1-5
Lignin 26-34 22-30
Extractives 0-5 0-10

 

153

 

Cellulose is a crystalline, high molecular weight polysaccharide and is a key
component of wood or any natural fiber. Cellulose consists of a long chain of beta
glucose. Due to the high presence of hydroxyl groups, cellulose has an affinity for water.
Microfibrils or microcrystals of cellulose are formed by hydrogen bonding of several
individual cellulose chains. These microfibrils are primarily responsible for the tensile

strength of the fiber [1-3].

Hemicellulose is similar in structure to cellulose but with low molecular weight
and is amorphous in nature. These characteristics make hemicellulose more prone to
hydrolysis than cellulose. Hemicellulose is composed of several heteropolymers such as
xylan, glucomannan, arabinose and others. Hemicellulose and cellulose are largely

responsible for the hydrophilic nature of wood [1-3].

Lignin is the second most abundant component of the wood. It acts as a binder in
wood and is mainly responsible for the strength of the wood. Lignin is a complex natural

cross-linked polymer whose definite structure is still not known [Ii-3].

Extractives account for 0-10% of the composition of wood, depending on the
species. Various extractives in the wood include oils, fats, resins, fatty acids, waxes,
tannins, etc. Extractives from the wood can be removed by a solvent extraction process if
needed. The type and amount of extractives can have a significant effect on the

manufacturing process and final properties of the composites [1-3].

Freshly cut wood contains around 50% by weight of moisture or water. Broadly,

the water in the wood can be classified into two forms, free and bound. The water in the

154

cell lumen is called free water as it is not chemically bonded to wood and can be easily
removed on drying. The water present in the cell walls is bound by hydrogen bonding
and therefore called bound water. Bound water needs extra energy in terms of heat for
removal [1, 2]. Dry wood can contain moisture ranging from 1 — 10 % by weight. It is
very important to dry the wood before using it for manufacturing composites. If not dried
properly the moisture in the wood is released during processing under high temperature
and pressure, causing several problems in both the manufacturing process and the end
product properties and quality. For example, moisture released during processing can
cause unwanted foaming and voids in the composite, reducing its mechanical properties

and aesthetics.

References:

1. Laboratory, F.P., Wood Handbook: Wood as an Engineering Material, Madison,
WI: Forest Products Society. 463 (1999).

2. Marra, A.A., Technology of Wood Bonding: Principles in Practice, New York:
Van Nostrand Reinhold. 454 (1992).

3. Haygreen, J .G. and Bowyer, J .L., Forest Products and Wood Science : An
Introduction. 3rd ed, Ames: Iowa State University Press. xv, 484 p. (1996).

155

 

Polyvinyl Chloride:

Polyvinyl chloride (PVC) is a synthetic polymer that consists of repeating units of
vinyl chloride monomer (Figure B.1). The preparation of this monomer was first
reported by Regnault in 1835. Baumann reported preparation of PVC polymer in a
sealed vacuum tube under sunlight in 1872 [1]. Today’s commercially available PVC is
polymerized by a free radical polymerization mechanism using suspension (80% of

production), emulsion (12%) or bulk/mass (8%) polymerizing techniques.

 

 

Figure A.1. Structure of PVC (where n = 700 — 1500)

PVC is mostly amorphous in nature, but has a small amount of crystallinity. It is
a polar polymer due to the presence of chlorine atoms. PVC has a glass transition
temperature of around 80°C. It has limited thermal stability and starts to decompose at as
low as 100°C, and tends to adhere to metallic surfaces when heated [2]. As a result, PVC
is extremely difficult to process alone and requires the incorporation of several additives
to ease the processing and to tailor the end product properties. Various additives used in
PVC processing include: stabilizers, lubricants, plasticizers, processing aids, impact
modifiers and fillers. Flexibility in formulating the PVC to achieve desired properties

gives it a popularity and acceptability for a wide range of products ranging from rigid

156

 

pipes for construction to flexible tubing for medical applications. On the other hand,
these additives complicate the recycling process of PVC. Since each product
manufactured from PVC contains different type and amount of additives, recycled PVC
contains an unknown mix of additives. This makes it difficult to obtain desired

properties in products made from recycled PVC.

References:

1. Brydson, J.A., Plastics Materials. 7th ed, Oxford ; Boston: Butterworth-
Heinemann. xxvii 920p. (1999).

2. Titow, W.V., PVC Technology. 4th ed, London ; New York: Elsevier Applied
Science Pub. xxx, 1233 p. (1984).

157

 

Polylactide:

Polylactide (PLA) is a renewable and biodegradable linear polyester (Figure C.l)
formed from lactic acid. PLA was first polymerized in 1920 but was not commercialized
until recently. NatureWorks LLC (previously known as Cargill Dow LLC) is a major
producer of PLA from corn starch in the US, producing 140,000 metric tons per year.
The general process of making lactic acid involves the fermentation of dextrose, which
can be made from several different sources such as biomass or corn. The presence of
chiral carbon atoms results in two different forms of lactic acid: L-form and D-forrn.
PLA can be polymerized by several different techniques, the most common of which is
ring-opening polymerization of lactide dimer. Polymerization of PLA is a stepwise
process. The first step involves the low molecular weight polymerization of lactic acid,
followed by depolymerization of this low molecular weight polylactic acid into lactide
dimers using a catalyst. Finally, these lactide dimers are further polymerized through

ring opening polymerization into PLA [1].

Lactide dimer exits in three different forms: L-lactide formed from two L-lactic
acid molecules, D-lactide formed from two D-lactic acid molecules and meso-lactide
formed from one each of L-lactic acid and D-lactic acid. Desired properties in the PLA
can be achieved by varying the ratio of different forms of the lactide. The polymer that
consists of more than 93% of L-form lactic acid is semicrystalline in nature, while a
lower content of the L-form (50 — 93%) results in an amorphous polymer [1].
Microorganisms used in fermentation of dextrose produce predominantly L-lactic acid.

As a result, most commercially produced PLA consists of L-form. However, some

158

 

quantity of meso-lactide is almost always present, which makes PLA a copolymer of L-

and meso-lactide.

H3C

 

O

Lactide

CH3

 

ring-opening _

 

polymerizationr

CH3

 

O

——(fH—C—o——CH—fi—o--—
0

H3

 

_n

Polylactide

Figure A.2. Polymerization of Polylactide

Another process to produce PLA is through direct condensation polymerization of

lactic acid. Although direct condensation polymerization is simple in theory, in practice

it is hard to get solvent-free high molecular weight polymer from this process.

Additional chemicals are needed to get high molecular weight polymer, which adds extra

cost to the overall process.

References:

1.

Auras, R., Harte, B., and Selke, S., "An Overview of Polylactides as Packaging
Materials", Macromolecular Bioscience, 4(9): 835-864 (2004).

159

 

APPENDIX B

Response: Die swell ratio

ANOVA for Selected Factorial Model
Analysis of variance table [Partial sum of squares]

Sum of Mean F
Source Squares DF Square Value Prob > F
Model 0.52 3 0.17 79.21 0.0005 significant
A 0.1 7 1 0.1 7 78. 00 0.0009
8 0.31 1 0.31 140.45 0. 0003
AB 0. 042 I 0.042 19.1 7 0.0119
Curvature 4.050E-003 l 4.050E-003 1.85 0.2458 not significant
Residual 8.775E-003 4 2.194E-003
Cor Total 0.53 8

The Model F-value of 79.21 implies the model is significant. There is only
a 0.05% chance that a "Model F-Value" this large could occur due to noise.

Values of "Prob > F " less than 0.0500 indicate model terms are significant.

In this case A, B, AB are significant model terms.

Values greater than 0.1000 indicate the model terms are not significant.

If there are many insignificant model terms (not counting those required to support hierarchy),
model reduction may improve your model.

The "Curvature F-value" of 1.85 implies the curvature (as measured by difference between the
average of the center points and the average of the factorial points) in the design space is not
significant relative to the noise. There is a 24.58% chance that a "Curvature F-value"

this large could occur due to noise.

Std. Dev. 0.047 R-Squared 0.9834
Mean 1.34 Adj R-Squared 0.9710
C.V. 3.51 Pred R-Squared N/A
PRESS N/A Adeq Precision 19.622

Case(s) with leverage of 1.0000: Pred R-Squared and PRESS statistic not defined

"Adeq Precision" measures the signal to noise ratio. A ratio greater than 4 is desirable. Your
ratio of 19.622 indicates an adequate signal. This model can be used to navigate the design
space.

Coefficient Standard 95% CI 95% CI
Factor Estimate DF Error Low High
VIF
Intercept 1.34 1 0.017 1.30 1.39
A-Wood flour -0.15 1 0.017 -0.19 -0. 10
1.00
B-K - 400 0.20 1 0.017 0.15 0.24
1.00
AB -0.072 1 0.017 -0.12 -0.027

160

 

1.00
Center Point -0.068 1 0.050 -0.21 0.070
1.00

Final Equation in Terms of Coded Factors:

Die swell =
+1.34
-0.15 * A
+0.20 * B

-0.072 * A "' B

Final Equation in Terms of Actual Factors:

 

Die swell = E
+1.22000 '
-2.95000E-003 * Wood flour
+0.067188 * K - 400
-7.25000E-004 * Wood flour * K - 400

Response: Viscosity (500)

Transform: Power Lambda: -0.5 Constant: 0
ANOVA for Selected Factorial Model
Analysis of variance table [Partial sum of squares]

Sum of Mean F
Source Squares D Square Value Prob > F
Model 1.181E-003 2.362E-004 651.74 0.0015 significant
A 2. 453E-004 2.453E-004 676. 79 0. 0015
B 7.366E-005 7.366E-005 203.22 0.0049

F
5
1
1
C 4.172E-004 1 4.172E-004 1151.06 0.0009
AB 2.593E-004 1 2.593E-004 715.45 0. 0014
AC 1.856E-004 I 1.856E-004 512.17 0.0019
Curvature 6.535E-006 l 6.535E-006 18.03 0.0512 not significant
Residual 7.249E-007 2 362513-007
Cor Total 1.188E-003 8
The Model F-value of 651.74 implies the model is significant. There is only
a 0.15% chance that a "Model F -Value" this large could occur due to noise.

Values of "Prob > F " less than 0.0500 indicate model terms are significant.

In this case A, B, C, AB, AC are significant model terms.

Values greater than 0.1000 indicate the model terms are not significant.

If there are many insignificant model terms (not counting those required to support hierarchy),
model reduction may improve your model.

The "Curvature F-value" of 18.03 implies there is curvature (as measured by difference between
the average of the center points and the average of the factorial points) in the design space.

161

There is only a 5.12% chance that a "Curvature F-value" this large could occur due to
noise.

Std. Dev. 6.020E-004 R-Squared 0.9994
Mean 0.068 Adj R-Squared 0.9979
C.V. 0.88 Pred R-Squared N/A
PRESS N/A Adeq Precision 78.224

Case(s) with leverage of 1.0000: Pred R-Squared and PRESS statistic not defined

"Adeq Precision" measures the signal to noise ratio. A ratio greater than 4 is desirable. Your
ratio of 78.224 indicates an adequate signal. This model can be used to navigate the design
space.

 

Coefficient Standard 95% CI 95% Cl
Factor Estimate DF Error Low High VIF
Intercept 0.068 1 2.129E-004 0.067 0.069 E
A-Wood flour-5.537E-003 l 2.129E-004 -6.453E-003 -4.622E-003 1.00
B-K - 400 -3.034E-OO3 l 2.129E-004 -3.950E-003 -2.119E-003 1.00
C-DOP 7.222E-003 l 2.129E-004 6.306E-003 8.137E-003 1.00
AB 5.693E-003 l 2.129E-004 4.778E-003 6.609E-OO3 1.00
AC -4.817E-003 l 2.129E-004 -5.733E-003 -3.901E-003 1.00
Center Point -2.711E-003 1 6.386E-004 -5.459E-003 3.620E-005 1.00

Final Equation in Terms of Coded Factors:

(Viscosity (500))‘0-5 -_—_

+0.068
-5.537E-003 * A
-3.034E-003 * B
+7.222E-003 "‘ C

+5.693E-003 "‘ A * B
-4.817E-003 * A * C

Final Equation in Terms of Actual Factors:

(Viscosity (500))-0-5 =
+0.070626
-2.56551E-004 "‘ Wood flour
-2.18196E-003 * K - 400
+6.0194OE-003 * DOP
+5.69345E-005 * Wood flour * K - 400
-9.63434E-005 * Wood flour * DOP

162

Response: Viscosity (1200)

Transform: Power Lambda: -1 Constant: 0
ANOVA for Selected Factorial Model

Analysis of variance table [Partial sum of squares]

Sum of Mean F
Source Squares DF Square Value Prob>F
Model 3.649E-005 5 7.298E-006 1675.20 0.0006 significant
A 8.563E-006 I 856313-006 1965.66 0.0005
B 1.232E-006 I 1.232E-006 282.84 0.0035
C 1.369E-005 l 1.369E-005 3143.20 0.0003
AB 6.320E-006 1 6.320E-006 1450.70 0.0007
AC 6.681E-006 1 6.681E-006 1533.58 0.0007
Curvature 6.950E-008 l 6.950E-008 15.95 0.0573 not significant
Residual 8.713E-009 2 4.357E-009
Cor Total 3.657E-005 8

The Model F-value of 1675.20 implies the model is significant. There is only
a 0.06% chance that a "Model F-Value" this large could occur due to noise.

Values of "Prob > F " less than 0.0500 indicate model terms are significant.

In this case A, B, C, AB, AC are significant model terms.

Values greater than 0.1000 indicate the model terms are not significant.

If there are many insignificant model terms (not counting those required to support hierarchy),
model reduction may improve your model.

The "Curvature F-value" of 15.95 implies there is curvature (as measured by difference between
the average of the center points and the average of the factorial points) in the design space.
There is only a 5.73% chance that a "Curvature F-value" this large could occur due to

noise.
Std. Dev. 6.600E-005 R-Squared 0.9998
Mean 6. l 96E-003 Adj R-Squared 0.9992
C.V. 1.07 Pred R-Squared N/A
PRESS N/A Adeq Precision 120.373

Case(s) with leverage of 1.0000: Pred R-Squared and PRESS statistic not defined

"Adeq Precision" measures the signal to noise ratio. A ratio greater than 4 is desirable. Your
ratio of 120.373 indicates an adequate signal. This model can be used to navigate the design
space.

Coefficient Standard 95% CI 95% CI
Factor Estimate DF Error Low High VIF
Intercept 622713-003 l 2.334E-005 6.126E-003 6.327E-003
A-Wood flour-1.035E-003 l 2.334E-005 -l.l35E-OO3 -9.342E-OO4 1.00
B-K -400 -3.925E-004 l 2.334E-005 -4.929E-004 -2.921E-OO4 1.00
C-DOP 1.308E-003 l 2.334E-005 1.208E-003 1.409E-OO3 1.00
AB 888813-004 l 2.334E-005 788413-004 9.892E-004 1.00
AC -9.139E-OO4 l 2.334E-005 -l.014E-003 -8.l35E-004 1.00
Center Point -2.796E-004 l 7.001E-005 -5.808E-004 2.160E-005 1.00

163

 

Final Equation in Terms of Coded Factors:

(Viscosity)'1
+6.227E-003
-l .035E-003
-3.925E-004
+1 .308E-003

*A
*B
*C

+8.888E-004 * A * B
-9.139E-004 * A * C

Final Equation in Terms of Actual Factors:

(Viscosity)'l
+6.32031E-003

-4.0383ZE-005 * Wood flour

-3.20321E-004 * K - 400
+1.11109E-003 "‘ DOP
+8.88821E-006 * Wood flour * K - 400
-l .82772E-005 * Wood flour * DOP

Response: Viscosity (2000)

Transform: Power Lambda:
ANOVA for Selected Factorial Model

Analysis of variance table [Partial sum of squares]

Sum of

Source Squares
Model 8.899E-007
A 2.277E-007
B 1.055E-008
C 3.554E-007
AB 1.094E-007
AC 1.869E-007
Curvature 8.69lE-011
Residual 1.025E-010
Cor Total 8.901E-007

DF

(1!

mN—‘NNNNN

-1 .5 Constant:

Mean F
Square Value
1.780E-007 3473 .22
2.27 7E-007 4442. 90
1.055E-008 205.95
3. 554E-007 6934.84
1. 094E-007 2134.27
18695-007 3648.12
8.69113-011 1.70
5.124E-01 l

0

Prob > F
0.0003 significant
0. 0002
0.0048
0.0001
0. 0005
0. 0003
0.3226 not significant

The Model F-value of 3473.22 implies the model is significant. There is only

a 0.03% chance that a "Model F-Value" this large could occur due to noise.

Values of "Prob > F" less than 0.0500 indicate model terms are significant.
In this case A, B, C, AB, AC are significant model terms.
Values greater than 0.1000 indicate the model terms are not significant.

If there are many insignificant model terms (not counting those required to support hierarchy),
model reduction may improve your model.

164

The "Curvature F-value" of 1.70 implies the curvature (as measured by difference between the
average of the center points and the average of the factorial points) in the design space is not
significant relative to the noise. There is a 32.26% chance that a "Curvature F-value"

this large could occur due to noise.

Std. Dev. 7. 159E-006 R-Squared 0.9999
Mean 6.754E-004 Adj R-Squared 0.9996
C.V. 1.06 Pred R-Squared N/A
PRESS N/A Adeq Precision 163.744

Case(s) with leverage of 1.0000: Pred R-Squared and PRESS statistic not defined

"Adeq Precision" measures the signal to noise ratio. A ratio greater than 4 is desirable. Your
ratio of 163.744 indicates an adequate signal. This model can be used to navigate the design
space.

Coefficient Standard 95% CI 95% CI
Factor Estimate DF Error Low High VIF
Intercept 6.765E-004 l 2.53lE-006 6.656E-004 6.874E-004
A-Wood flour-1.687E-OO4 1 2.53lE-OO6 -1.796E-004 -1.578E-004 1.00
B-K -400 -3.632E-005 l 2.53lE-006 -4.721E-005 -2.543E-005 1.00
C-DOP 2.108E-OO4 1 2.531E-006 1.999E-004 2.217E-004 1.00
AB 1.169E-004 l 2.531E-006 1.06OE-004 1.278E-004 1.00
AC -l.529E-004 l 2.53lE-OO6 -l.638E-004 -l.420E-004 1.00
Center Point 988813-006 l 7.593E-006 -4.256E-005 2.278E-005 1.00

Final Equation in Terms of Coded Factors:

(Viscosity (2000))-1 .5 =

+6.765E-004

-l .687E-004 * A
-3.632E-005 * B
+2.108E-004 "' C

+1.169E-004 *A’B
-l.529E-004 * A * C

Final Equation in Terms of Actual Factors:

(Viscosity (2000))-1-5 =
+6.34845E-004
-5.31025E-006 * Wood flour
-3.83l l4E-005 * K - 400
+1.81816E-004 * DOP
+1.16924E-006 * Wood flour "' K - 400
-3.05735E-006 * Wood flour * DOP

165

Table 8.1. Mechanical Properties of the PVC/wood-flour Composites made with
different concentration of coupling agent compared with the control sample.

 

 

 

 

 

 

Samples Flexural Strength Flexural Modulus
MPa MPa
PVC 52.96 i 2.97 2168.6 i 74.9
Composite 44.51 i 2.09 3086.0 1; 146.6
Chitin'
2.5 wt. % 49.04 i 2.78 3521.9 3: 148.2
5 wt. % 52.39 i 3.27 3379.1 : 236.9
6.67 wt. % 54.23 i 4.65 3667.5 i 232.8
7.5 wt. % 54.09 i 2.10 3545.5 j; 251.7
10 wt. % 53.79 j; 1.92 3557.3 1 171.9
Chitosanl
0.5 wt. % 53.05 i 3.86 3497.1 1 244.5
1.0 wt. % 50.27 i 2.49 3282.9 : 219.2
1.5 wt. % 51.75 i 2.53 3457.1 i 221.7
2.5 wt. % 50.24 i 1.66 3375.1 1 245.0

 

lchitin and chitosan wt. % were based on the weight of the wood flour.

166

Table B.2. Mechanical Properties of the PLA/wood-flour Composites made with
different concentration of coupling agent.

 

 

 

3;; Wag/51w Chm-1‘ 8523;: 5132:3131
(MPa) (MPa)

100 o 0 99.64 : 1.6 3180.3 :608

80 20 0 97.97 :07 4249.3 :138.4 "
70 30 0 97.40 :1.9 5020.6 :111.1
60 40 0 96.13 :41 5953.1 :2108 _.
60 40 1 96.91 :14 6107.9 :151
60 40 2.5 97.74 :24 5945 :122

60 40 5 98.37 :25 5919.4 :1385

60 40 10 99.06 :25 6040.6 :172.6

 

I Chitosan content is based on the weight of wood flour

167

 

1111111111111111111811111110