$333?"

{.111 e

. ”thIVuMEr
“ﬂaw hm. h. «Hun
.ﬁ..». "5%.. .w
:éakén .u. h.
. 24W}

54". .
._ a...
2:.

L51: 2923.
rhﬂuw:§
2...».
1.5.
’v‘

 

 

 

 

 

 

LlBRARY

M‘Chigan State
University

 

This is to certify that the
dissertation entitled

Improving the Durability and Mechanical Properties of Wood-
Plastic Composites Through Coextrusion

presented by

Shan Jin

has been accepted towards fulﬁllment
of the requirements for the

PhD. degree in Forestry

 

 

‘ C A
MARK? aq’iudet

Major Professor’s Signature
Twig an ¥ c1010

Date

 

MSU is an Afﬁnnative Action/Equal Opportunity Institution

 

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5108 K:IProj/Aco&Pres/ClRCIDateDue.indd

 

IMPROVING THE DURABILITY AND MECHANICAL PROPERTIES OF
WOOD-PLASTIC COMPOSITES THROUGH COEXTRUSION

By

, Shan Jin

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Forestry

2010

ABSTRACT

IMPROVING THE DURABILITY AND MECHANICAL PROPERTIES OF
WOOD-PLASTIC COMPOSITES THROUGH COEXTRUSION

By
Shan Jin

Wood-plastic composites (WPCs) can absorb moisture in a humid
environment due to the hydrophilic nature of the wood in the composites, making
products susceptible to microbial growth and loss of mechanical properties. This
study tested the concept of using coextrusion technology to manufacture two-layer
coextruded wood-plastic composites encapsulated with a plastic surface-rich cap layer
for improving their water resistance, mechanical performance, and weathering
resistance. The results indicated that WPCs can be encapsulated by a neat rigid PVC
or HDPE cap layer, or a PVC-based composites (wood ﬂour or carbon nanotube ﬁlled
rigid PVC) cap layer in a coextrusion process. The moisture uptake rate was lower
when a PVC surface-rich cap layer was applied in the composites, and the extent of
the decrease was a strong function of the amount of wood ﬂour in the cap layer but
insensitive to cap layer thickness. Coextruding PVC surface-rich cap layers on WPCs
signiﬁcantly increased the ﬂexural strength but decreased the ﬂexural modulus as
compared with those of control samples. The changes in bending properties were
sensitive to both cap layer thickness and wood ﬂour content in the cap. In order to
obtain coextruded WPCs with better ﬂexural properties than uncapped WPCS, a
two-level factorial design was used to evaluate the statistical effects of material
compositions and processing conditions on these properties. Material composition
variables were the wood ﬂour content in the core layer and the carbon nanotube (CNT)

content in the cap layer of coextruded composites. The processing condition variable

was the processing temperature proﬁle for the core layer. Factors leading to a fast
fusion of the PVC-wood ﬂour composites in the core layer, i.e. low wood ﬂour
content and high processing temperature, improved the ﬂexural properties of
coextruded composites. Reinforcing the cap layer with CNTs produced a signiﬁcant
improvement in the ﬂexural properties of the coextruded composites, insensitive to
the core layer composition and the processing temperature condition. Moreover, this
study also examined the effect of coextruding a clear HDPE cap layer onto
HDPE/wood-ﬂour composites on the discoloration of coextruded composites exposed
to accelerated UV weathering tests. Two separate discoloration characteristics
occurred in the discoloration of composites. For uncapped WPCs, chemical changes
due to photooxidation resulted in darkening followed by physical changes, including
loss of colored wood components from the surface, as well as increased roughness on
the surface, which led to lightening of WPCs. By contrast, because a hydrophobic cap
layer prevented the loss of colored components from the surface, coextruding a clear
hydrophobic HDPE cap layer over WPCs signiﬁcantly decreased the discoloration
upon weathering. Photooxidation of wood components at the interface accounted for
the discoloration of coextruded WPCs before the failure of cap layer. As the cap layer
absorbed a speciﬁed amount of UV light and reduced oxygen available to interface, it
decreased the photooxidation rate at the interface compared to that at the WPCs

surface.

Copyright by
ShanJht

2010

Dedicated to my parents

WEI yueming and JIN jiazhen

ACKNOWLEDGEMENTS

First and foremost, I am grateful to my advisor, Dr. Laurent Matuana, for giving
me this great opportunity to further my education in the ﬁeld of composites materials,
and for all of his support, guidance, and encouragement. I also thank my other committee
members, Dr. Donatien Pascal Kamdem, Dr. Pascal Nzokou, and Dr. Maria Rubino.

Their advice helped me expand my understanding of research.

Assistance with FTIR from Dr. Ruth Smith, Dr. Rui Huang, and Dr. Faai Zhang
in the Department of Chemistry as well as UVN IS from Kelby Thayer in the School of
Packaging is appreciated. I am thankful for the following staffs in the Department of
Forestry for their assistance throughout my graduate program: Juli Kerr, Carol
Graysmith, Barb Anderson, and Karen Ellerhorst. I am also grateful to all my colleagues
(Dr. Omar Faruk, Carlos Diaz, Kojo Afrifah, etc.) in the Biocomposites Manufacturing

Laboratory for their continued help and support.

This work was ﬁnancially supported by the USDA—CSREES Grant-Advanced
Technology Applications to Eastern Hardwood Utilization (Grant No: 2006-34158-17581
and 2008-34158-19510). An additional thank you to the companies listed in the text of

chapters 3, 4, and 5 for generously donating the materials used in the research.

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................... x

LIST OF FIGURES .......................................................................................................... xii

LIST OF ABBREVIATIONS ............................................................................................ xv
CHAPTER 1

Introduction .......................................................................................................................... 1

1.1 Introduction ........................................................................................................ l

1.2 Hypotheses ......................................................................................................... 5

1.3 Objectives .......................................................................................................... 6

1.4 Structure of dissertation ..................................................................................... 7

1.5 References .......................................................................................................... 8
CHAPTER 2

Background and Literature Review ................................................................................... 12

2.1 Polymer ............................................................................................................ 12

2.1.1 General Properties ................................................................. 12

2.1.2 Polyethylene .......................................................................... 15

2.1.3 Poly(vinyl chloride) .............................................................. 16

2.1.4 Photodegradation of PE and PVC ......................................... 20

2.2.5 Protection against photodegradation ..................................... 25

2.2 Wood ................................................................................................................ 31

2.2.1 Chemical composition of wood ............................................ 31

2.2.2 Anatomic structure of wood .................................................. 37

2.2.3 Moisture absorption of wood ................................................ 39

2.2.4 Photodegradation of wood .................................................... 44

2.2.5 Protection against photodegradation ..................................... 50

2.3 Wood-plastic composites ................................................................................. 52

2.3.1 Introduction ........................................................................... 52

2.3.2 Effect of wood content .......................................................... 52

2.3.3 Effect of the characteristics of the wood ﬁber ...................... 52

2.3.4 Effect of the characteristics of the polymer matrix ............... 53

2.3.5 Effect of surface modiﬁcation of wood and the addition of

coupling agents .............................................................................. 54

2.3.6 Effect of manufacturing methods .......................................... 55

2.3.7 Moisture absorption of WPCs ............................................... 55

2.3.8 Photodegradation of WPCs ................................................... 56

2.3.9 Protection against photodegradation ..................................... 57

2.4 Testing and property evaluation ....................................................................... 58

2.4.1 Weathering testing ................................................................ 58

2.4.2 X-ray photoelectron spectroscopy ........................................ 60

vii

2.4.3 Fourier transform infrared spectroscopy ............................... 64

2.4.4 Color measurements .............................................................. 65

2.4.5 F lexural properties ................................................................ 67

2.5 References ........................................................................................................ 68
CHAPTER 3

Coextruded PVC/Wood-Flour Composites with WPC Cap Layers .................................. 74

3.1 Abstract ............................................................................................................ 75

3.2 Introduction ...................................................................................................... 76

3.3 Experimental .................................................................................................... 79

3.3.1 Materials ............................................................................... 79

3.3.2 Compounding and coextrusion of Composites ..................... 79

3.3.3 Property evaluation ............................................................... 84

3.4 Results and Discussion .................................................................................... 90

3.4.1 Color measurements .............................................................. 90

3.4.2 Moisture absorption .............................................................. 93

3.4.3 Flexural properties ................................................................ 97

3.5 Conclusions .................................................................................................... 101

3.6 Acknowledgements ........................................................................................ 102

3.6 References ...................................................................................................... 103
CHAPTER 4

Wood Plastic Composites Coextruded with Multi-walled Carbon Nanotube Filled Rigid

PVC Cap Layer ................................................................................................................ 105

4.1 Abstract .......................................................................................................... 106

4.2 Introduction .................................................................................................... 107

4.3 Experimental .................................................................................................. 111

4.3.1 Materials ............................................................................. 111

4.3.2 Rigid PVC formulation ....................................................... 113

4.3.3 Compounds for core layer ................................................... 113

4.3.4 Compounds for outer layer ................................................. 113

4.3.5 Experimental design ............................................................ 114

4.3.6 Coextrusion of PVC/wood-ﬂour composites ...................... 117

4.3.7 Flexural property evaluation ............................................... 120

4.3.8 Fusion tests .......................................................................... 120

4.3.9 Scanning electron microscopy ............................................ 121

4.4 Results and Discussion .................................................................................. 122

4.4.1 Statistical signiﬁcance of effects ........................................ 122

4.4.2 Effect of the interaction factor on the ﬂexural strength ...... 133

4.4.3 Effect of the interaction factors on the ﬂexural modulus ...142

4.5 Conclusions .................................................................................................... 148

4.6 Acknowledgements ........................................................................................ 150

4.7 References ...................................................................................................... 151

viii

CHAPTERS
Ultraviolet Weathering of HDPE/Wood-Flour Composites Coextruded with a Clear

HDPE Cap Layer ............................................................................................................. 155

5.1 Abstract .......................................................................................................... 156

5.2 Introduction .................................................................................................... 157

5.3 Experimental .................................................................................................. 165

5.3.1 Materials ............................................................................. 165

5.3.2 Compounding and extrusion ............................................... 165

5.3.3 Accelerated weathering tests ............................................... 168

5.3.4 Scanning electron microscopy (SEM) ................................ 168

5.3.5 UV-vis reﬂectance spectroscopy ........................................ 168

5.3.6 Fourier transform infrared spectroscopy ............................. 169

5.3.7 X-ray photoelectron spectroscopy ...................................... 169

5.3.8 Color measurements ............................................................ 170

5.4 Results and Discussion .................................................................................. 171

5.4.1 Morphology and surface characteristics of weather

composites .................................................................................... 171

5.4.2 Color Analysis .................................................................... 184

5.4.3 Discoloration mechanisms of WPCs .................................. 192

5.5 Conclusions .................................................................................................... 194

5.6 Acknowledgements ........................................................................................ 195

5.7 References ...................................................................................................... 196
CHAPTER 6

Conclusions and Future Work ......................................................................................... 200

6.1 Conclusions .................................................................................................... 200

6.2 Future work .................................................................................................... 203

ix

Table 2.1.

Table 2.2.

Table 3.1.

Table 3.2.

Table 3.3.

Table 3.4.

Table 4.1.

Table 4.2.

Table 4.3.

Table 4.4.

Table 4.5.

Table 4.6.

Table 4.7.

LIST OF TABLES

Equilibrium moisture content of wood at various temperature and relative
humidity. .................................................................................................... 41

Bond dissociation energies and radiation wavelength ............................... 45

Formulations used in core and cap layers of coextruded rigid PVC/wood-
ﬂour composites. ........................................................................................ 80

Effects of the composition and processing speed of cap layer on its
thickness, surface color, and moisture absorption of coextruded rigid

PVC/wood-ﬂour composites and control samples ..................................... 85

Effect of the thickness of cap layer on the ﬂexural properties of

coextruded WPCs ....................................................................................... 98
Effects of wood ﬂour content in cap layers on the ﬂexural properties of

coextruded WPCs ....................................................................................... 99
Formulations of rigid PVC ....................................................................... 112

Factorial design matrix in terms of both actual and coded factor levels. 116
Processing conditions for coextrusion process ........................................ 119

ANOVA for selected factorial model of ﬂexural strength of coextruded
composites ................................................................................................ 130

ANOVA for selected factorial model of ﬂexural modulus of coextruded
composites ................................................................................................ 13 1

Effects of wood ﬂour content and test temperature on the fusion time of
rigid PVC/wood-ﬂour composites. .......................................................... 138

Comparison of ﬂexural properties between uncapped (control) and
coextruded rigid PVC/wood-ﬂour composites. Composites contained 50%
wood ﬂour. ............................................................................................... 147

Table 5.1.

Table 5.2.

Table 5.3.

Table 5.4.

Assignment of absorption bands of IR spectrum ..................................... 176

Concentrations of oxidative products on uncapped and
coextruded WPCs ..................................................................................... 180

C13 peak analysis at HDPE, wood ﬁber, and WPCs surface and at
coextruded WPCs interface determined by high-resolution XPS scan....183

Color parameters of control and coextruded samples before and after
weathering. ............................................................................................... l 86

xi

Figure 2.1.
Figure 2.2.
Figure 2.3.
Figure 2.4.
Figure 2.5.

Figure 2.6.

Figure 2.7.
Figure 2.8.

Figure 2.9.

Figure 2.10.
Figure 2.11.
Figure 2.12.
Figure 2.13.
Figure 2.14.
Figure 2.15.
Figure 2.16.
Figure 2.17.
Figure 2.18.

Figure 2.19.

Figure 3.].

LIST OF FIGURES
Images in this thesis/dissertation are presented in color.
Repeating structural units of polymer ....................................................... 14
Fusion curve of neat rigid PVC compound ............................................... 19

Photodegradation of polymers initiated by photolysis of hydroperoxides 22

Photodegradation of polymers initiated by photolysis of ketone .............. 24
Protection mechanisms of benzotriazoles and benzophenones ................. 27
UV absorption spectra of HPT, BTZ, and BP ........................................... 28
The stabilization mechanism of HALS ..................................................... 30
The structure of cellulose .......................................................................... 32
The structure of lignin precursors .............................................................. 34
The structure of lignin precursors .............................................................. 35
The layered cell wall structure of softwood .............................................. 38
Moisture sorption isotherms of wood at 70 0F. ......................................... 42
The IUPAC classiﬁcation for adsorption isotherms .................................. 43
Sequence of photooxidation reactions in wood ......................................... 46
Reﬂection spectra of wood, lignin, and cellulose ...................................... 48
The sunlight spectrum ............................................................................... 59
A survey spectrum of HDPE-based WPCs ............................................... 61
A high resolution spectrum of HDPE-based WPCs in C13 region ............ 63
Color space (CIE L*a*b* (CIELAB) ) ...................................................... 66
Schematic of the coextrusion system ........................................................ 82

xii

Figure 3.2.

Figure 3.3.

Figure 3.4.

Figure 4.1.

Figure 4.2.

Figure 4.3.

Figure 4.4.

Figure 4.5.

Figure 4.6.

Figure 4.7.

Cross section of unﬁlled rigid PVC cap layer coextruded on WPCs. The
arrows show the areas of poor interfacial adhesion between cap and core
layers .......................................................................................................... 86

WPC samples coextruded with unﬁlled PVC cap layer, 5% wood ﬂour
ﬁlled PVC cap layer, 20% wood ﬂour ﬁlled PVC cap layer, and control
WPC (no cap layer) .................................................................................... 91

Moisture absorption curves of control samples and coextruded rigid
PVC/wood-ﬂour composites with various types of cap layers: wood-ﬂour
(5-20 wt%) ﬁlled rigid PVC cap layers and unﬁlled rigid PVC cap layers
of various thicknesses. ............................................................................... 94

Schematic of the coextrusion system ....................................................... 118

Normal probability plots of effects for (a) ﬂexural strength and (b)
modulus of coextruded rigid PVC/wood-ﬂour composites. .................... 123

Perturbation plots of the ﬂexural strength of coextruded composites
against wood ﬂour content in core layer and CNT content in cap layer of
coextruded composites processed at (a) low temperature and (b) high
temperature proﬁle for the core layer. ..................................................... 125

Perturbation plots of the ﬂexural modulus of coextruded composites
against wood ﬂour content in core layer and CNT content in cap layer of
coextruded composites processed at (a) low temperature and (b) high
temperature proﬁle for the core layer. ..................................................... 127

Interaction graphs of the variation in ﬂexural strength as a function of the
interaction between wood ﬂour content in the core and CNT content in the
cap layers of coextruded composites processed at (a) low temperature and
(b) high temperature proﬁle for the core layer ......................................... 134

SEM micrographs of the rigid PVC/wood-ﬂour composites core layer
containing various amounts of wood ﬂour and processed with different
temperature proﬁles: (a) 40% wood ﬂour/high temperature; (b) 40% wood
ﬂour/low temperature; (c) 60% wood ﬂour/high temperature; (d) 60%
wood ﬂour/low temperature ..................................................................... 140

Interaction graphs of the variation in ﬂexural modulus of coextruded
composites as a function of the processing temperature for core layer-
wood flour content in inner layer interaction: (a) no CNT in cap layer and
(b) cap layer reinforced with 5% CNT. ................................................... 143

)(iii

Figure 5.1.

Figure 5.2.

Figure 5.3.

Figure 5.4.

Figure 5.5.

Figure 5.6.

Figure 5.7.

Figure 5.8.

Figure 5.9.

Schematic of the cross section of coextruded HDPE/wood-ﬂour
composite ................................................................................................. 167

SEM (X8Ok) of surface of noncoextruded WPCs (left column) and
coextruded WPCs (right column) as a function of UV radiation exposure
times: (a and t) 0 h, (b and g) 432 h, (c and h) 744 h, (d and i) 1392 h, and
(eandj)1952h ........................................................................................ 172

UV transmission spectra of ﬁlms of HDPE” (for cap layer) and HDPEOA

(polymer matrix for uncapped WPC and core layer of coextruded
WPCs) ..................................................................................................... 174

F TIR spectra of wood ﬂour (maple) and HDPE (a) before and (b) after
exposure to 1952 h UV ............................................................................ 175

FTIR of the surface of uncapped WPCs as a function of weathering time
(a) unexposed or control, (b) 192 h, (c) 432 h, (d) 744 h, (e) 1392 h, and
(f) 1952 h ................................................................................................. 178

FT IR of the interface of coextruded WPCs as a function of weathering
time (a) unexposed or control, (b) 192 h, (c) 432 h, (d) 744 h, (e) 1392 h,
and (f) 1952 h .......................................................................................... 179

Visual appearance of uncapped WPCs (left column) and coextruded
WPCs (right column) as a function of weathering exposure times (0, 192,

432, 744, 1392, 1952 11) ........................................................................... 185

Color parameters for solid wood (maple), HDPE0_4 and HDPELZ ......... 188

Micrograph of HDPE cap layer in “white” area (a) and “brown” area (b) at

the surface of weathered coextruded samples ......................................... 191
xiv

BP
BTZ
CNT
EMC
FTIR

HALS
HDPE
HPT
LDPE
LLDPE
MC
MFI
MOR
MOE
MUR

PE

phr

PVC

SEM

UV

LIST OF ABBREVIATIONS

Hydroxybenzophenone
Hydroxyphenylbenzotriazole

Carbon Nanotube

Equilibrium Moisture Content

Fourier Transform Infrared Spectrometer
Hindered Amine Light Stabilizers

High Density Polyethylene
Hydroxyphenyl-s-triazine

Low Density Polyethylene

Linear Low Density Polyethylene
Moisture Content

Melt Flow Index

Modulus of Rupture (strength)

Modulus of Elasticity (stiffness)
Moisture Uptake Rate

Polyethylene (high-density or low-density)
Parts Per Hundred Parts of Resin
Polyvinyl Chloride

Relative Humidity

Scanning Electron Microscopy

Ultraviolet (rays or light)

XV

UVA -- UV Absorbers
WPCs -- Wood-Plastic Composites

XPS -- X-ray Photoelectron Spectroscopy

xvi

CHAPTER 1

Introduction

1.1 Introduction

Wood-plastic composites (WPCs) are manufactured by mixing wood ﬁber and
plastics through plastic processing procedures such as compression, extrusion or injection
molding [1-5]. Wood residue and recycled plastics can be used to produce WPCs, saving

waste disposals and reducing the raw material cost [6-7].

Wood-plastic composites combine the excellent properties of both wood and
plastics. They are stiffer and more environmentally friendly compared to unﬁlled plastic
owing to the greater stiffness and degradability of wood ﬁbers compared to plastics [4-5].
Wood-plastic composites also require less maintenance than solid wood due to their

hydrophobic polymer phase [8].

Due to these attributes, wood-plastic composites have experienced signiﬁcant
market expansion in recent years as a replacement for solid wood, mainly in outdoor
applications such as railings, decking, fences, and window and door frames [9-10]. These
products are marketed as low-maintenance and weather-resistant compared to solid wood,
because the wood ﬁber in WPCs can be encapsulated in a hydrophobic polymer matrix
[ll-12]. Moreover, most of current WPC manufacturers are offering 15 to 20 years of

warranty on their products. However, as WPCs are exposed to an outdoor environment of

moisture and ultraviolet light (UV), the expectation that the products would maintain

their new appearance without maintenance is not realistic [13].

When in ground contact, WPCs can be damaged by biological agents such as
fungi and termites. Microbial degradation results in surface stains and loss in both weight
and mechanical properties [14-17]. In an aboveground outdoor environment, moisture
and UV result in degradation of WPCs [18-23]. Moisture absorption reduces the adhesion
between the wood ﬁber and the polymer matrix, leading to loss of mechanical properties
[18-19]. Photodegradation can result in breakdown of both wood ﬁber and polymer

chains, causing loss in appearance and mechanical properties [20-23].

In addition, water absorption is a critical factor in both microbial degradation and
photodegradation. Wood decay fungi require wood moisture contents in excess of ﬁber
saturation to propagate, thus moisture content of 25-30% is necessary for fungal decay
[14]. In a photodegradation process, the presence of water accelerates oxidation reactions,
removes degraded components, and swells the wood ﬁber facilitating light penetration,
all of which would accelerate color fading and/or mechanical properties loss [24].
Therefore, improving the water resistance of WPCs is key to enhance the durability. It is
necessary to reduce or delay the moisture uptake of WPCs to prolong their long term

performance.

Many researchers have investigated the moisture/water absorption behaviors of
natural ﬁber—plastic composites. Marcovich et a1. investigated moisture diffusion in
polyester-(wood ﬂour) composites and found that the equilibrium moisture content of the
composites was dependent on the size and shape of the wood ﬂour [25]. Stark studied the

inﬂuence of moisture absorption on the mechanical properties of (wood ﬂour)-

polypropylene composites and reported that incomplete encapsulation of wood ﬂour by
the polypropylene matrix occurs in both 20% and 40% wood ﬂour composites [18]. In
addition, she demonstrated that more moisture was absorbed by the composites with
higher wood ﬂour loading, a situation that resulted in a signiﬁcant decrease of both
ﬂexural and tensile properties. Similar results were reported by Bledzki and Faruk, who
found that the impact properties of (wood ﬁberypolypropylene composites were affected
by moisture content [19]. Hygrotherrnal aging behavior of injection-molded (rice husk)-
ﬁlled polypropylene composites also has been examined by immersing the specimens in
distilled water at 30°C and 90°C [26]. It was found that both the tensile strength and the
modulus deteriorated after immersion in water, the extent of the deterioration being
dependent on the immersion temperature. This reduction in tensile properties was

attributed to the interfacial degradation in the composites.

Although several authors have investigated the moisture/water absorption,
biological degradation and photodegradation behaviors of (natural ﬁber)—plastic
composites, only a few provided the methods to alleviate these problems. Current
approaches to improving the durability of WPCs focus on the bulk of WPCs, i.e.,
incorporation of additives into the entire product. For example, adding coupling agents
improves the adhesion between the wood ﬁber and the polymer matrix [19]; adding
photostabilizers and pigments provides UV protection [24, 27-28]; adding biocides
prevents fungi and termite attacks [29]. However, the weathering effect primarily occurs
at the material surface [20-24, 30]. Adding preservatives to the bulk of material is an

expensive way to cope with weathering. WPCs with weathering protection on the surface

only will be a cost-effective way to improve durability. There are several methods for

providing a surface that can withstand weathering.

Coating is one of the methods widely used to improve the durability of wood
composites [31-33]. However, due to low surface energy of plastic it is difﬁcult to coat
the plastic components at the surface of WPCs. Also, coating involves solvent
evaporation, which often causes environmental concerns. Furthermore, drying stages

require extra time and reduce efﬁciency of assembly line.

Lamination is another method for providing a protective surface. Through melt
fusion or an adhesive layer, a surface layer can be laminated over substrate. But, when
those laminated composites are bent, delamination often occurs [34]. Moreover, it is
difﬁcult to use lamination technology to manufacture a fully encapsulated structure other

than a planar layered structure.

Coextrusion is another method for providing a protective surface since
coextrusion technology can produce a multilayered product with different properties at
outer and inner layers, thus offering different properties between surface and bulk [3 5-36].
The beneﬁt of coextrusion is that each layer of coextruded material offers a particular
property, including impermeability or resistance to moisture uptake and improved
mechanical properties, such as stiffness, depending on the compositions of the outer and
inner layers. For example, when composites are bent, one side suffers maximum
compressive stress, and the other side suffers maximum tensile stress [37]. Therefore, the
composition of the outer layer will inﬂuence not only the durability, but also the ﬂexural

properties of the overall composites.

Coextrusion technology is a promising method to manufacture WPCs with a
durable surface since the additives involved in durability improvement will only be
incorporated into the outer layer of WPCs rather than into the entire product. As a result,
the durability of WPCs product could be improved in a cost-effective way. Similarly,
ﬂexural properties of coextruded products are related to the surface property of materials.
Thus, if the surface layer is reinforced with nanoﬁllers such as nanoclays or carbon
nanotubes [38-39], the ﬂexural properties of the composites would be increased

accordingly.

Although the coextrusion process has been extensively studied for polymers to
manufacture multilayer ﬁlms, sheet, and tubing [40-42], there is no published work on
the use of this innovative technology for the development of wood-plastic composites
with enhanced mechanical performance and durability, which is the main goal of this

research.

1.2 Hypotheses
In this project, we plan to test the following hypotheses:

l. Coextruding a plastic rich outer layer on WPCs is an effective way to improve the

durability of WPCs in terms of water/vapor and weathering resistance;

2. Coextruding a reinforced outer layer on WPCs will enhance the ﬂexural

properties of WPCs.

1.3 Objectives

This project tests the idea of using coextrusion technology to make two-layer

coextruded encapsulated WPCs for improving their mechanical performance and

durability in a cost-effective way. The following speciﬁc objectives will be addressed:

a.

Investigate the effects of cap layer compositions (cap layer type and wood ﬂour
content) and processing conditions (affecting cap layer thickness) on the
water/vapor absorption behaviors of both noncoextruded WPCs (control) and
coextruded WPCs with the ultimate goal of identifying the moisture resistant

WPC formulations;

Analyze the effects of cap layer formulation (unﬁlled vs. carbon nanotube ﬁlled),
core layer formulation (wood ﬂour content) and melt rheology (high and low
temperature proﬁle for the core layer) on the ﬂexural properties of both control
and coextruded WPCs in order to statistically model and optimize the material

compositions-processing conditions-properties relationships;

Investigate the effects of cap layer types (noncoextruded vs. coextruded) and
weathering conditions on the UV degradation of both control and coextruded
WPCs to understand the role of the cap layer in the mechanisms involved in

weathering of WPCs.

Coextruded WPCs that will improve the durability (in terms of resistance to
moisture absorption and weathering) and mechanical performance of WPCs in an
effective way is expected to result from this project. Moreover, reduced cost of
weathering protection and enhanced performance will spur the application of WPCs and

utilization of wood from forest products industry.

1.4 Structure of Dissertation

The ﬁrst chapter introduces the rationale of this research. A broad background on
wood-plastic composites was reviewed in the second chapter. The inﬂuence of the cap
layer types on the ﬂexural properties and water resistance of WPCs was discussed in
Chapter 3. The material composition-processing condition-ﬂexural property relationships
for coextruded WPCs were analyzed and modeled in Chapter 4. Chapter 5 examined the
ultraviolet weathering of HDPE/wood-ﬂour composites coextruded with a clear HDPE
cap layer in order to gain an in-depth understanding of the role of transparent HDPE cap
layer on the UV resistance of WPCs. Finally, Chapter 6 provides the overall conclusions
and future work on the role of coextruding a plastic rich cap layer on the durability of

WPCs.

1.5 References

10.

ll.

12.

13.

14.

CM. Clemons, Wood-plastic composites in the United States: the interfacing of
two industries, Forest Products Journal, 52(6): 10-18 (2002).

AK. Bledzki, J. Gassan, and S. Theis, Wood-ﬁlled thermoplastic composites,
Mechanics of Composite Materials, 34(6): 563-568 (1998).

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

M. Bengtsson and K. Oksman, Silane crosslinked wood plastic composites:
Processing and properties, Composites Science and Technology, 66(13): 2177-2186
(2006)

A. Ashori, Wood—plastic composites as promising green-composites for automotive
industries, Bioresource Technology, 99(11): 4661-4667 (2008).

S.K. Najaﬁ, A. Kiaefar, and E. Hamidina, Water absorption behavior of composites
from sawdust and recycled Plastics, Journal of Reinforced Plastics and Composites,
26(3): 341-348 (2007).

Y Lei, Q. Wu, F Yao, and Y. Xu, Preparation and properties of recycled
HDPE/natural ﬁber composites, Composites Part A: Applied Science and
Manufacturing, 38(7): 1664-1674 (2007).

JJ. Morrell, N.M. Stark, D.E. Pendleton, and A.G. McDonald, Durability of wood-
plastic composites, Wood Design Focus, 16(3): 7-10 (2006).

A. Jacob, WPC industry focuses on performance and cost, Reinforced Plastics,
50(5): 32-33 (2006).

J. Markarian, Outdoor living space drives growth in wood-plastic composites,
Reinforced Plastics, 10(4): 20-25 (2008).

J.M. Pilarski and L.M. Matuana, Durability of wood ﬂour-plastic composites
exposed to accelerated freeze-thaw cycling. Part 1. Rigid PVC matrix, Journal of
Vinyl & Additive Technology, 11(1): 1-8 (2005).

J.M. Pilarski and L.M. Matuana, Durability of wood ﬂour-plastic composites
exposed to accelerated freeze-thaw cycling. Part II. High density polyethylene
matrix, Journal of Applied Polymer Science, 100(35): 35-39 (2006).

J. Markarian, Wood-plastic composites need protection, Plastics Engineering, 63(7):
54-56 (2007).

CM. Clemons and ER. Ibach, Application of laboratory fungal resistance tests to
solid wood and wood-plastic composite, Plastics in Building Construction, 27(2):
7-14 (2003).

15.

16.

17.

18.

I9.

20.

21.

22.

23.

24.

25.

26.

A. Schirp and MP. Wolcott, Inﬂuence of fungal decay and moisture absorption on
mechanical properties of extruded wood-plastic composites, Wood and Fiber
Science, 37(4): 643-652 (2005).

B. Dawson-Andoh, L.M. Matuana, and J. Harrison, Susceptibility of high density
polyethylene/wood-ﬂour composite to mold discoloration, Journal of the Institute
ofWood Science, 17(2): 114-119 (2005).

B. Dawson-Andoh, L.M. Matuana, and J. Harrison, Mold susceptibility of rigid
PVC/wood-ﬂour composites, Journal of Vinyl & Additive Technology, 10(4): 179-
186 (2004).

N.M. Stark, Inﬂuence of moisture absorption on mechanical properties of wood
ﬂour-polypropylene composites, Journal of Thermoplastic Composite Materials,
14(5): 421-432 (2001).

AK. Bledzki and O. Faruk, Creep and impact properties of wood ﬁbre—
polypropylene composites: Inﬂuence of temperature and moisture content,
Composites Science and Technology, 64(5): 693-700 (2004).

L.M. Matuana and DP. Kamdem, Accelerated ultraviolet weathering of
PVC/wood-ﬂour composites, Polymer Engineering and Science, 42(8): 1657-1666
(2002)

N.M. Stark and L.M. Matuana, Surface chemistry and mechanical property changes
of wood-ﬂour/high-density-polyethylene composites after accelerated weathering,
Journal of Applied Polymer Science, 94(6): 2263-2273 (2004).

N.M. Stark and L.M. Matuana, Surface chemistry changes of weathered
HDPE/wood-ﬂour composites studied by XPS and FTIR spectroscopy, Polymer
Degradation and Stability, 86(1): 1-9 (2004).

N.M. Stark and L.M. Matuana, Characterization of weathered wood—plastic
composite surfaces using FTIR spectroscopy, contact angle, and XPS, Polymer
Degradation and Stability, 92(10): 1883-1890 (2007).

N.M. Stark and L.M. Matuana. Inﬂuence of photostabilizers on wood ﬂour-HDPE
composites exposed to xenon-arc radiation with and without water spray, Polymer
Degradation and Stability, 91 (12): 3048-3056 (2006).

NE. Marcovich, M.M. Reboredo, and MI Aranguren, Moisture diffusion in
polyester—woodﬂour composites, Polymer, 40(26): 7313—7320 (1999).

Z.A. M. Ishak, B.N. Yow, B.L. Ng, H.P.S.A. Khalil, and H.D. Rozman,
Hygrothermal aging and tensile behavior of injection-molded rice husk-ﬁlled
polypropylene composites, Journal of Applied Polymer Science, 81(3): 742-753
(2001)

27.

28.

29.

30.
31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

NM. Stark and L.M. Matuana, Ultraviolet weathering of photostabilized wood-
ﬂour-ﬁlled high density polyethylene composites, Journal of Applied Polymer
Science, 90(10): 2609-2617 (2003).

M. Muasher and M. Sain, The efﬁcacy of photostabilizers on the color change of
wood ﬁlled plastic composites, Polymer Degradation and Stability, 91(5): 1156-
1165(2006)

D.E. Pendleton, T.A. Hoffard, T. Adcock, B. Woodward, and MP. Wolcott,
Durability of extruded HDPE/wood composites, Forest Products Journal, 52(6):
21-27 (2002).

D.N. -S Hon, Wood and Cellulose Chemistry, Marcel Dekker, New York, (2001).

ST. Chang and PL. Chou, Photo-discoloration of UV-curable acrylic coatings and
the underlying wood, Polymer Degradation and Stability, 63 (3): 435-439 (1999).

ST. Chang and PL. Chou, Photodiscoloration inhibition of wood coated with UV-
curable acrylic clear coatings and its elucidation, Polymer Degradation and

‘ Stability, 69(3): 355-360 (2000).

PL. Chou, H.T. Chang, T.F. Yeh, and ST. Chang, Characterizing the conservation
effect of clear coatings on photodegradation of wood, Bioresource Technology,
99(5): 1073-1079 (2008).

L. Long, M.P. Wolcott, J. Zhang, and K. Englund, Flexural properties of surface
reinforced wood/plastic deck board, Polymer Engineering and Science, 47(3): 281-
288 (2007).

C. Rauwendaal, Understand Extrusion, Hanser/Gardner, Cincinnati, (1998).

H.J. Giles Jr., J.R. Wagner Jr., and EM. Mount III, Extrusion: The Definitive
Processing Guide and Handbook, William Andrew Publishing, New York, (2004).

Z. Gﬁrdal, R.T. Haftka, and P. Hajela, Design and Optimization of Laminated
Composite Materials, John Wiley & Sons, New York, (1999).

SC. Tjong. Structural and mechanical properties of polymer nanocomposites,
Materials Science and Engineering R-Reports, 53(3-4): 73-197 (2006).

O. Faruk and L.M. Matuana, Reinforcement of rigid PVC/wood—ﬂour composites
with multi-walled carbon nanotubes, Journal of Vinyl and Additive Technology,
14(2): 60-64 (2008).

M. Panico, L.D. Maio, V. Nicolais, and L. Incamato, Experimental set-up and
characterization of a mLLDPE-PP co-extruded bilayer system, Polymer
Engineering and Science, 46(7): 874-881 (2006).

10

41.

42.

43.

44.

45.

P. Dole, L. Ave’rous, C. Joly, G.D. Valle, and C. Bliard, Evaluation of starch-PE
multilayers: processing and properties, Polymer Engineering and Science, 45(2):
217-224 (2005).

S. Elkoun, M. A. Huneault, K. McCormick, F. Puterbaugh and L. Kale, LLDPE-
based mono- and multilayer blown ﬁlms: property enhancement through
coextrusion, Polymer Engineering and Science, 45(9): 1222-1230 (2005).

L.M. Matuana and Q. Li, Statistical modeling and response surface optimization of
extruded HDPE/wood-ﬂour composite foams, Journal of Thermoplastic Composite
Materials, 17(2): 185-199 (2004).

J.W. Kim, K. Carlborn, L.M. Matuana, and RA. Heiden, Thermoplastic
modiﬁcation of urea-formaldehyde wood adhesives for increased moisture
resistance, Journal of Applied Polymer Science, 101(6): 4222-4229 (2006).

L.M. Matuana and J .W. Kim, Fusion characteristics of wood-ﬂour ﬁlled rigid PVC
by torque rheometry, Journal of Vinyl & Additive Technology, 13(1): 7-13 (2007).

11

CHAPTER 2

Background and Literature Review

WPCs is manufactured by mixing wood and plastics similar to plastic processing
procedures such as compression, extrusion, or injection molding [1-4].Wood refers to
various kinds of natural ﬁbers in different amounts and geometries (ﬁber, ﬂour, etc.)

whereas plastic is mainly thermoplastics.

2.1 Polymer

2.1.1 General properties

Polymers are large molecules (macromolecules) composed of repeating structural
units connected by covalent chemical bonds to form long chains. The composition of a
repeat unit, the number of repeat units, the spatial arrangement of repeat units on the long
chains, and chain linearity determine physical, chemical, and rheological properties of

polymers.

Polymers fall into two categories: therrnoset and thermoplastic. The manufacture
of therrnoset materials employs curing, a cross-linking process. In this process, the
molecular chains react at chemically active sites, linking into a rigid, 3-D structure [5].
This results in a melting point of material higher than the decomposition temperature.
Therefore, after curing, melting and remolding of a thermoset material cannot occur. In

contrast, therrnoplastics have no limit to repeated melting and solidifying through heating

l2

and cooling. Thermoplastics instead of thermosets are primarily used as polymer matrices
for WPCs. Thermoplastics mostly used in the manufacturing of WPCs include

polyethylene (PE), polypropylene (PP), and polyvinyl chloride (PVC) (Figure 2.1).

13

ﬂ—r'p —=
E—O --’JU
5 l

Figure 2.1. Repeating structural units of a polymer. R=H for PE; R=CH3 for PP; R=Cl

for PVC.

14

About 600 million pounds of thermoplastics in 2005 were used for the
manufacturing of WPCs in North America [4]. PE accounts for about 90% of the volume,
and PP and PVC occupy the remaining of 10% [4]. Polyethylene and Polyvinyl chloride

will be described in detail because they are the two main polymers used in this study.

2.1.2 Polyethylene

Polyethylene (PE) is the most popular plastic in the world [6]. PE consists of long
chains of the monomer ethylene. The mechanical properties of PE largely depend on
variables such as the extent and type of branching, the crystal structure, and the molecular
weight. Classiﬁcation of polyethylene is mainly based on its density and branching. The
main categories are high density polyethylene (HDPE), low density polyethylene (LDPE),
and linear low density polyethylene (LLDPE). HDPE typically has a density in the range
of 0.94-0.97 g/cm3 [6]. HDPE has a low degree of branching, leading to stronger
intermolecular forces and strength, and a higher degree of crystallinity. The densities of
LDPE are between 0.91-0.94 g/cm3 [6]. LDPE has a high degree of short and long chain
branching responsible for the weak intermolecular force and a lower degree of
crystallinity, which induces a lower tensile strength and higher ductility compared to
those of HDPE. The densities of LLDPE are in the range of 0915-0925 g/cm3 [7].
LLDPE, a substantially linear polymer, has a pronounced amount of short branches but
does not have long chain branches. Therefore, properties of LLDPE such as tensile and

ﬂexural properties are often between those of HDPE and LDPE.

HDPE is the major category of PE used as a polymer matrix in WPCs because

there are several advantages of HDPE over LDPE and LLDPE in terms of WPCs

15

application. HDPE has a higher degree of crystallinity than that of LDPE and LLDPE,
thus the properties of HDPE related to crystallinity such as modulus, strength, hardness,

and creep resistance override those of LDPE and LLDPE [4].

Besides density, another criterion to differentiate the polymers is the melt ﬂow
index (MFI), which is a measure of the ease of ﬂow of the melt of a thermoplastic
polymer [5]. It is deﬁned as the rate of extrusion of molten polymers through a die of a
speciﬁed length and diameter under prescribed conditions of temperature, load, and
piston position in the barrel. MFI depends on average molecular weight, molecular
weight distribution, and the degree of chain branching. It provides a critical consideration
in terms of material selection for manufacturing WPCs. Generally, injection molding
needs higher MFI polymers, while extrusion processes can perform with lower MFI

polymers [8].

2.1.3 Poly(vinyl chloride)

Polyvinyl chloride (PVC) is the third widely used thermoplastic polymer, aﬁer

polyethylene and polypropylene.

PVC resin is a major component in PVC formulation among other ingredients.
The K-value of PVC resin, related to the average number of molecules, reﬂects the
inherent viscosity. The K-value of PVC resin affects the processing and properties of the
ﬁnal product. In general, a lower K-value means lower melt viscosity and easier
processing at a given temperature, while a higher K-value relates to better mechanical
properties [9]. As a result, the properties of PVC-based WPCs depend on the K-value of

PVC resin in PVC formulation.

16

PVC resin requires additional additives, such as plasticizers, heat stabilizers and
processing aids. Plasticizers decrease the interaction among the PVC chains and thus
decrease the strength and modulus of PVC compounds. Flexible PVC compounds contain
plasticizers, while rigid PVC compounds do not contain plasticizers. Thus rigid PVC has
superior strength and modulus than ﬂexible PVC, which make rigid PVC preferable as

polymer matrix in WPCs.

Heat stabilizers are added to PVC formulation to improve thermal stability at
processing temperatures. Processing aids are added to change the melt-processing
behaviors of PVC such as fusion, melt strength, and melt viscosity. Lubricants are added
to loosen the internal friction between materials and external friction between materials

and processing equipment. Impact modiﬁers are added to improve the toughness [9].

The suspension polymerization process is the principal commercial route to
produce PVC resins in industry. Thus PVC resin is granular in nature. PVC resin is
composed of powder particles that are irregular in shape and about 100-200 pm in
diameter. These PVC powder particles are composed of microparticles that are on the
order of 10 to 15 microns in diameter. The primary particle agglomerates are composed

of submicroparticles, which have diameters of 1 micron [9, 10].

Melting of PVC compounds, generally referred to as PVC fusion, is the thermal
reduction of resin particle boundaries by heat and shearing and the compound approaches
a uniﬁed homogeneous structure. The procedure of PVC fusion includes the following
steps: The ﬁrst step is a breakdown of the powder particles and part of the microparticles.
Then grain boundaries must be eliminated, and the microparticles must be compacted

together. After further interdiffusion of PVC polymer chains, the boundaries of

17

submicroparticles disappear, and polymer is in a molten state [11, 12]. Since fusion is
involved in the breakdown of the particles and the interdiffusion of the polymer chains,

shearing force and energy are required.

Fusion behavior is described by fusion curve shown in F igure.2.2. Fusion curve
illustrates the changes of viscosity related to torque, temperature, and torque versus time.

The viscosity-related torque curve shows three different points.

When the PVC compound is loaded into the system, the ﬁrst point A is generated.
Then, the torque begins to decrease sharply, and this generates the second point B
because of free material ﬂow before the materials begin to compact. After that, the torque
begins to increase, and this generates the third point. At this point, X, the material reaches
a void-free state and ‘starts to melt at the interface between the compacted material and
the hot metal surface [12]. The time spent between the loading point (A) and the fusion

point (X) is deﬁned as fusion time.

18

N 00 45- 01 O)
O O O O O
l 1 I 1 1

-'O— ——;:
(“ﬂ

Torque (Nm)

A
O
1

 

I F I T I

0 50 100 150 200 250 300
Time (s)

0

Figure 2.2. Fusion curve of neat rigid PVC compound [12].

Fusion behavior is a linkage between processing conditions and end-use
properties. A lot of research shows an inadequate fusion could lead to poor properties of
PVC compounds. It has been reported that the tensile strength of rigid PVC increased
with increasing fusion and then leveled off at the higher levels of fusion [10]. The effect
of fusion level on the brittle-ductile behavior and stiffness of rigid PVC has also been
investigated. While the brittle-ductile behavior was found to achieve an optimum value

and then decrease, the stiffness increased with fusion level in the whole range [10].

2.1.4 Photodegradation of PE and PVC

Photodegradation is a degradation of a molecule caused by the absorption of
photons, which are from ultraviolet (UV) radiation, visible light, and infrared radiation
(IR) [13-16]. The photodegradation rate of polymer relies on sunlight intensity,
temperature, humidity, oxygen pressure, and the type of polymer [l3]. Photodegradation
cannot occur without chromophores which absorb photo-energy [14]. Although PE does
not contain chromophores inherently, introduction of chromophores such as
hydroperoxide and carbonyl groups occurs during polymer manufacturing, processing
and storage [13-16]. On the other hand, the photodegradation of PVC is due to the
unsaturation site (conjugated double bond), low dissociation energy of C-Cl linkage, and
hydroperoxide groups and carbonyl groups introduced during manufacturing, processing

and storage [13].

There are two major mechanisms responsible for the photodegradation of polymer.
One is initiated by photolysis of hydroperoxides and occurs in three stages, as shown in

Figure 2.3 [13]. The ﬁrst step is the initiation of free radicals generated by photolysis of

20

hydroperoxides. The second step, propagation, contains free polymer alkyl radicals
reacting with oxygen and producing the polymer peroxy radicals whereupon the energy
of polymer peroxy radicals can transfer to another polymer, generating secondary
polymer radicals. Polymer peroxy radicals undergo various kinds of reactions, which
results in chain scission. The third step, termination, incorporates two free radicals

combining together, thereby leading to crosslinking.

21

hv
Initiation: POOH -> P-, PO-, H0, H0;-

Propagation: P- + 02 —+ P020
P02°+ PH —-* P- + POOH
POz-——> -CHO + CH2=CH- + -OH
Termination: P02- + POz- —-> P=O + P-OH + 02
P020 + HOz- —> P=O + HzO + 02 or POOH + 0;

H02“ + HOz' —* HOOH + 02

Figure 2.3. Photodegradation of polymers initiated by photolysis of hydroperoxides

(PH: polymer) [13].

22

The other photodegradation mechanism of polymer is initiated by photolysis of ketone [13]. The
photodegradation proceeds through Norrish I and II reactions, as shown in Figure 2.4. Both
Norrish I and 11 reactions result in the scission of polymer backbone. In Norrish 1 reaction, a
carbon-carbon bond is broken to form one alkyl free radical and one carbonyl free radical. As a
result, carbon monoxide is easily lost from carbonyl free radical. Thus, two alkyl free radicals are
generated via Norrish I. These radicals readily react with oxygen to generate alkyl peroxy
radicals and propagate the reaction [13]. Free radical reaction may terminate via crosslinking. In
Norrish II reaction, a carbon-carbon bond is broken. Methyl ketone and alkene are generated.
Therefore, both Norrish I and 11 reactions result in the scission of polymer backbone. Norrish I

also can leads to crosslinking.

23

.7. 3 : 5802 SW Ea H szoZ A3 .888. mo Embouosm 3 BEES E0833 mo cosgﬁwouoaonm .vd oSwE

:51 01 All :5“ ul

__ _
0 IO

Nno“ ol + ”no” :01 ll 1:51 £01 clamol ”no 1
_ Z. __
:o o
e

co + 1:5 All .0 l EDI

__
O

l :5 + .o l :51 l 1 £01 blame!
__ .5 __
o o
E

24

Although both chain scission and crosslinking occurs as the result of
photodegradation, chain scission is the dominant mechanism in polyethylene and PVC
[17]. Chain scission decreases molecular weight, leading to increase the chain mobility
and surface cracks [17—20]. Consequently. crystallinity increases but impact resistance
and elongation decrease [18]. Increasing crystallinity also affects optical properties, as
the crystalline region is translucent and the amorphous region is transparent [5]. Thus

polyethylene became whiten after photodegradation.

2.1.5 Protection against photodegradation

UV absorbers (UVA) and hindered amine light stabilizers (HALS) are primarily
light stabilizers for Polyoleﬁn and PVC [21-25]. The corresponding protection
mechanisms prevent the initial reaction and delay the propagation reaction by

consumption of the free radicals.

The protection mechanism of UV absorbers is based on the absorption of the UV
radiation and dissipation of radiation as heat. The UV absorption follows the Lambert-

Beer’s law [21, 22]:

logil— = -8 - l -c (Equation 2.1)
0

Where I0 represents the light energy of monochromatic light incident per unit time on a

substrate with a single absorbing species of concentration c (mole per L), I the energy

Past through the substrate. 1 the thickness, and a the molar extinction coefﬁcient of the

25

absorbing species. Since UV absorption by UVA is thickness dependent, UVA may fail

to provide adequate surface protection for thin portion.

Hydroxybenzophenone (BP), hydroxyphenylbenzotriazole (BTZ), and
hydroxyphenyl-s-triazine (HPT) are three main classes of UVA commercially used in
polymers and coatings [22, 23]. The typical protection mechanism of benzotriazoles and
benzophenones are illustrated in the Figure 2.5. Hydrogen bonding in the o-hydroxy
groups pronouncedly inﬂuences photochemical properties [22]. Different classes of UVA
have different UV absorption spectra, as shown in Figure 2.6. While BP mainly provides
protection against UVB radiation (280-320 nm), HPT and BTZ provide protection against

both UVB and UVA radiation (320-400 nm).

Pigments absorb or reﬂect UV radiation and can be considered as another type of
UV stabilizer. The major difference between UVA and pigments is that while UVA
primarily absorb radiation in the UV region, pigments absorb light in both UV and VIS
area. Thus materials often change color or decrease transparency after the addition of
pigments. Titanium dioxide, carbon black, and zinc oxide are pigments used frequently as

UV stabilizers.

26

(a)

 

 

 

HO R1 T 0 R1
Q“): -—‘-- t“
N
m' x, O
R2 R2
(b)
H
H
O O 0/ 0
Light \
0 C O 0
OR OR
Figure 2.5. Protection mechanisms of (a) benzotriazoles and (b) benzophenones [22].

27

 

1.2

 

 

A —- UVAl n

1.0 / \

0.8 \ - . : UVA2 _
UVA2

M ./ _ _

 

 

 

0.6 / ' " w
0.4 A 5

UVA3 —\ _ \
0.2

290 310 330 350 370 390
Wavelength (nm)

 

Absorbance

 

 

 

 

Figure 2.6. UV absorption spectra of HPT (UVAI), BTZ (UVA2), and BP (UVA3)

[23].

28

Hindered amine light stabilizers (HALS), derivatives of 2,2,6,6-tetramethyl
piperidine, protect the polymer from UV radiation via free radical termination. The
stabilization mechanism of HALS is believed to be a cyclic chain-breaking antioxidant
process, shown in Figure 2.7. In the ﬁrst step, in the presence of oxygen and radiation,
the HALS compound (1) is converted into the corresponding nitroxyl radical (2), then
nitroxyl radical traps a free radical to form an alkoxy amine (NOR) (3). The alkoxy
amine reacts with a peroxide radical under formation of the intermediate structures (4)
which then decompose into harmless alcohols and ketones. At the same time, the nitroxyl
radical (2) is regenerated, which can repeat the cycle [24, 25]. Moreover, since HALS,
unlike UVA, is independent on sample thickness, HALS can be used in applications with
high surface-to-volume ratios such as ﬁlms, tapes, and ﬁbers [24]. Polymers often require

a lower addition level of HALS than UV absorbers [22, 24].

In order to achieve optimum UV protection, various types of UV stabilizers are
often used together [22, 24]. The selection of light stabilizers is dependent on the
characteristics of the polymer matrix (especially its absorption spectrum), application,

and dimension etc.

29

1 N-Fl — -—-—- 2 N-o’ 3 N-OR

R’OH + 3:0

>- {1 4....

 

 

Figure 2.7. The stabilization mechanism of HALS [25].

30

2.2 Wood

2.2.1 Chemical composition of wood

Cellulose, lignin, hemicelluloses, and minor amounts of extraneous materials (ash
and extractives) are the main components of dry wood. Each of these components

contributes to the properties of wood and wood derived products.

Cellulose

Cellulose, the primary component, constitutes approximately 40-50% of wood
substance on an oven-dry basis [26]. As a high-molecular-weight, stereoregular, and
linear polymer, cellulose consists of chains of repeating (l,4)-B-D-g1ucopyranose units
(Figure 2.8). With the empirical formula for cellulose represented as (C6H1005)n, the
degree of polymerization (DP) ranges from 1000 to 10000. Hydroxyl groups in the
cellulose molecules are the cause of water sorption of wood. Through intra- and
intermolecular hydrogen bonds, cellulose molecules crystallize. Cellulose is highly
crystalline and may contain up to 65% crystalline regions [27]. The surfaces of crystalline
cellulose are accessible to moisture sorption and chemical modiﬁcation, but the bulk of
the crystalline cellulose is not accessible. In contrast, most of the amorphous cellulose is
accessible to moisture sorption and chemical modiﬁcation except for regions covered by

hemicelluloses and lignin.

31

 

Figure 2.8. The structure of cellulose [27].

32

Hemicelluloses

Hemicelluloses account for approximately 20% of wood substance. While
cellulose is mainly composed of glucose units, hemicelluloses contain various kinds of
sugar units with a lower DP (100-200) than cellulose. Moreover, while cellulose is highly

crystalline and resistant to hydrolysis, hemicelluloses are amorphous and easily

hydrolyzed [27].

Lignin

Lignin, a three-dimensional phenylpropanol amorphous polymer, accounts for 18
to 25% of the wood substance in hardwoods and 25 to 33% in softwoods [27]. The
precursors of lignin biosynthesis are paracoumaryl alcohol, coniferyl alcohol, and sinapyl
alcohol (Figure 2.9). Paracoumaryl alcohol is a minor precursor of both softwood and
hardwood lignins. Coniferyl alcohol is the dominant precursor of softwood lignins. Both
coniferyl alcohol and sinapyl alcohols are precursors of hardwood lignin. These three
basic units are linked by ether and carbon to carbon linkages, forming lignin. The
structure of lignin is shown in Figure 2.10. It is relatively hydrophobic compared to
holocellulose (cellulose and hemicelluloses). Although lignin occurs in wood throughout
the cell wall, concentrations are highest towards the outside of and between the cells.
Lignin commonly acts as the cementing agent responsible for binding individual cells

together [26].

33

.1 20H 20H

 

/ /
0CH3
on OH
(1) (2) (3)

Figure 2.9. The structure of lignin precursors: (1) paracoumaryl alcohol, (2) coniferyl

alcohol, and (3) sinapyl alcohol [27].

34

 

H15 OH

 

 

 

 

 

tit-o-
econ
Hzcou
lcuzoul . H'-O—
Flt: -5 "3:0 j
"C --CH agree
I l
.. 0
© HJCO HCOH
a; H40" °“
0——CH HzEOH OCH,
neon Hc—o
HCOH . '
HZmI-l H0-,-( if
°—"" ago® or", or,
our,
HOEHIE HOE": "It. 0
"f 001, ”f‘”—_'
Hc—-o Hcou
o—pt Ht——o
HCOH (:0 [Al
H3CMOCH3
OH onto-c1

 

 

 

Figure 2.10. The structure of lignin in the soft wood [28].

35

Extractives and ash

Extractives, accounting for 1-2% of wood substance, consists of a wide range of
organic compounds mainly including fats, waxes, terpene, phenolic compounds, and
certain carbohydrates [27]. The quantities and qualities of extractives assist in
determining the wood’s permeability to liquids and inﬂuence other wood properties such
as density, hardness, and compressive strength. Extractives provide resistance to insect or
fungal attack in certain woods. Additionally, extractives generate the odors and colors
attributed to most woods [27]. The extractives are often classiﬁed by the solvent used for
their extraction. The di-chloromethane extractable content of wood refers to the measure
of components such as waxes, fats, resins, photosterols, and non-volatile hydrocarbons
[27]. The ethanol-benzene extractable content of the wood contains components such as
low molecular weight carbohydrates, salts, and other water-soluble substances. The

amount of extractives substantially depends on the seasoning or drying of wood [27].

Ash accounts for less than 1% of wood substance. Inorganic minerals, primarily

calcium, potassium, and magnesium, form ash.

36

 

2.2.2 Anatomic structure of wood

There are two domains in most cells of wood: the cell wall and the cell lumen [27].
The cell lumen is a void space. The cell wall is layer which surrounds the cell lumen. The
cell walls of wood are multilayer. They consist of the primary wall (P), and the layers of
the outer (SI), middle (S2), and inner (S3) secondary walls [26, 27]. The adjacent cells
are glued together by the middle lamella (m.l.). The layered cell wall structure of
softwood is shown in Figure 2.11. Difference between these layers lies in their structures,
orientations and number of ﬁbrils or ﬁbers, and chemical composition. Cellulose is
primarily located in the secondary wall. Lignin is distributed throughout the cell wall but
more in the middle lamella region. In contrast, hemicellulose is distributed in both the

cell wall and middle lamella.

37

 

 

 

 

Figure 2.11. The layered cell wall structure of softwood [27].

38

2.2.3 Moisture absorption of wood

Since holocellulose (cellulose and hemicelluloses) is rich in hydroxyl group,
wood absorb water easily and quickly. The moisture content (MC) of wood is calculated

by Equation 2.2:

MC = initial weight — oven dry weight

 

. (Equation 2.2)
oven dry weight

Water in wood can be categorized into free water (liquid water or water vapor in
cell lumens), and bound water (water held chemically within cell walls) [1]. The Fiber
saturation point (FSP) refers to the point at which all liquid water in the lumens has been
removed but the cell wall is still saturated. When moisture content is below the F SP,
physical and mechanical properties of wood change as moisture content changes. In
contrast, when moisture content is above the F SP, these properties of wood remain
constant as moisture content changes. The moisture content of wood depends on the
relative humidity (RH) and temperature of the environment. Equilibrium moisture
content (EMC) arises when the wood neither gains nor loses moisture at a given
temperature and RH. A strong relationship exists between EMC, relative humidity, and
temperature, as shown in Table 2.1 [26]. For most practical purposes, the values in Table
2.1 may be applied to wood of any species [26]. At given temperature, the relationship
between the equilibrium moisture content of a material and relative humidity can be
displayed graphically by a curve (Figure 2.12.), which is called moisture sorption
isotherm. Generally, there are six types of sorption isotherms according to the
International Union of Pure and Applied Chemistry (IUPAC) classiﬁcation [29]. The

moisture sorption isotherm of wood best ﬁts type II isotherm as shown in Figure 2.13,

39

which indicates strong adsorbate-adsorbent interactions [29]. In other words, interactions
between moisture and wood are strong. This is attributed to the formation of hydrogen

bonding between moisture and hydroxyl groups in the wood.

40

Table 2.1. Equilibrium moisture content of wood at various temperature and relative

humidity [26].

 

 

 

 

 

 

Temperature Moisture content (%) at various relative humidity
(°F) 10% 30% 50% 70% 95%
30 2.6 6.3 9.5 13.5 24.3
50 2.6 6.3 9.5 13.4 24.3
70 2.5 6.2 9.2 13.1 23.9
90 2.3 5.9 8.9 12.6 23.3
120 2.1 5.4 8.2 11.7 22.0
150 1.8 4.8 7.4 10.6 20.4

 

41

Moisture Content (%)

 

0 20 40 60 80 100
Relative Humidity (%)

Figure.2.12. Moisture sorption isotherms of wood at 70 °F. Data adapted from Table

2.1 [26].

42

 

 

IV V VI

Speciﬁc amount adsorbed n

 

 

 

 

 

Relative pressure p/p°

F igure.2.13. The IUPAC classiﬁcation for adsorption isotherms [29].

43

2.2.4 Photodegradation of wood

As mentioned in the section of photodegradation of polymer, photodegradation
cannot occur without chromophores which absorb photo-energy [14]. Moreover, a
reaction occurs only once a molecule obtains the necessary activation energy [30-32].
The bond dissociation energies of several functional groups in wood and corresponding
wavelengths are listed in Table 2.2 [27]. Because visible light in the range of 400-700 nm
carries energy less than 70 kcal/mol, it fails to provide sufﬁcient energy to break
chemical bonds in most of the wood components [30]. Therefore, the ultraviolet part of
sunlight (295-400 nm) predominantly accounts for the photodegradation of wood. The
general mechanism of photooxidation of wood is similar to that of polymer, including the

three same steps: initiation, propagation, and termination, shown in Figure 2.14 [31].

The breakdown of chemical bond by absorption of UV photons generates free
radicals. Free radicals react with oxygen and water, forming hydroperoxides. Carbon
radicals in a vacuum tend to stabilize themselves by recombination and hydrogen
abstraction. In contrast, in the presence of oxygen they are transformed rapidly into
hydroperoxide radicals [30]. Both free radicals and hydroperoxides can initiate chain

scission reactions.

44

Table 2.2. Bond dissociation energies and radiation wavelength [27].

 

Bond Dissociation

 

Wavelength
Bond Energy

(Kcal/mol) (nm)
C-C (Aromatic) 124 231
C-H (Aromatic) 103 278
C-H (Methane) 102 280
O-H (Methanol) 100 286
C-0 (Ethanol) 92 311
C-O (Methanol) 89 321
CH3COO-C (Methyl ester) 86 333
C-C (Ethane) 84 340
C-Cl (Methyl ether) 82 349
C-COCH3 (Acetone) 79 362
C-0 (Methyl ether) 76 376

 

45

hv
Initiation: RH -) R' + H'

Propagation: R' + 02 -> R02.
R02' +RH -> ROOH + R'
Termination: ROZ' +R02° -> nonradical product

R' +R02° -i nonradical product

R' +R' -> nonradical product

Figure 2.14. Sequence of photooxidation reactions in wood (RH:

hemicelluloses, and lignin) [31].

46

cellulose,

The degrees of photodegradation of wood components largely depend on their
ability to absorb UV light. The reﬂection spectra of wood, cellulose, and lignin are shown
in Figure 2.15. Wood absorbs both UV and the shorter wavelengths of visible light.
Lignin absorbs light in the both UV and visible regions more efﬁciently than cellulose
[33], due to the chromophoric groups in lignin (Figure 2.10). As a result, lignin
experiences more severe photodegradation than cellulose. The chemical component
analysis showed an increase in cellulose content and a decrease in lignin content at the

wood surface after weathering [3 0].

Absorption of UV light by wood initiates the photochemical reactions at surface.
When cellulose is exposed to sunlight, depolymerzatoin at the glycosidic linkanges,
dehydrogenation at the C1 and C5 positions, and dehydroxymethylation at the C5-C6
side chain take place [30]. When liginin is exposed to sunlight, guaiacyl and phenoxy
radicals are generated. The following demethylation of guaiacyl and phenoxy radicals
yields quinoid chromophores [34]. Quinones can cause an intense color change
(yellowing or darkening) even at very low concentration [35]. The photodegradation of
lignin leads to the decay of aromatic structures and the formation of quinone and
conjugated carbonyl groups [34]. Since the decrease of aromatic skeletal vibrations in the
1600 or 1510 cm'1 region and the increase of carbonyl group in the 1730 cm'1 region are
more pronounced than conjugated carbonyl bonds, the former is often used to evaluate

the decay of lignin instead of the latter [34-3 8].

47

Absorbance

 

 

 

 

1 L
0.5 -
0 . . -
250 300 3 50 400 450
Wavelength (nm)

Figure 2.15. Reﬂection spectra of (a) wood, (b) lignin, and (C) 6611111036 133]-

48

Study showed that removal of water/ethanol/toluene extractable compounds from
wood can signiﬁcantly change the color, especially the lightness [38]. However, after
expo sure to long-term weathering (2400b accelerate artiﬁcial weathering), both extractive
free and unextracted wood showed similar color in terms of lightness, redness, and
yellowness [39]. This can be attributed to the removal of degraded color compound from
surface by water spray. Since most of the extractive components are considered phenolic
in nature, their photodegradation may be similar to that of lignin, which is also phenolic
in nature. Investigation have shown that the FTIR spectra of extractive free wood
exposed to xenon arc irradiation experienced less reduction at 1510 cm‘1 peak compared
to unextracted wood [3 8], implying that extractives like lignin also experienced the decay

of benzyl ring.

The degraded products of weathered woods, are mainly organic acids, vanillin,
SYl‘ingaldehyde, and higher molecular weight compounds, which are all leachable [30].
Usually, wood in the outdoor environment initially become yellow and brown due to the
PhOtooxidation of lignin, and when the colored degraded products leach out, a silver-gray
layer composed of leach-resistant parts of the partially degraded cellulose remains on the

wood surface [30].

The intensity of UV/vis light decreases after penetrating wood deeper than 75-200
11m, making photodegradation of wood a surface phenomenon. However, the degradation
layer of wood can become deeper with the weathering. Kataoka et al. reported that the

degradation layer of sugi sapwood increased to 500 um after 1500 h of xenon arc

radiation [40].

49

Besides the intensity of UV/vis light, the rate of a photochemical process depends
on the moisture content at the wood’s surface, the oxygen concentration, the temperature
and the type of wood species [41, 42]. Water has various chemical and physical effects
on photodegradation. Liquid water can result in the erosion of oxidized layers, swelling
of matrix, and extraction of additives soluble in water [31]. Liquid or gaseous water also
may induce the hydrolysis of oxidation intermediates or of ﬁnal products [31]. Especially
for wood, water penetrates and swells the wood ﬁber, opening up non-accessible regions
for light penetration [30]. Oxygen is involved in the propagation and termination. Since
oxidation needs energy, there is temperature dependence for photodegradation. Wood

species dependence can be attributed to variety in chemical constituents of wood species.

Chemical changes also bring about physical changes. Weathering leads to
cracking and splitting of the wood, both on a microscopic level inside the cell walls and
as visible cracks on the surface [27, 30]. Consequently, weathered wood loses weight and

strength [30, 43-45].

22-5 Protection against photodegradation

There are two common methods to protect wood against photodegradation. One is

WOOd surface treatment and another is coating [43-46].

Wood surface treatment primarily refers to modiﬁcation of chemical composition
Of Wood surface by reaction. Grafting of UV absorber (UVA) to wood reduced both
Weight loss and tensile strength loss since UV absorber can reduce the UV light absorbed
by Wood [44]. High temperature or microwave irradiation is often needed to bond UVA

to Wood. Esteriﬁcation of wood with benzoyl choride also can decrease both the weight

50

loss and color change since benzoyl group can act as a UV screen protecting cellulose

and lignin [45, 46].

Coating also can provide protection against environmental damage such as
photodegradation, biodegradation, and mechanical abrasion [43, 47-49]. For protection
against photodegradation, opaque pigmented coatings block the UV/visible light but
cover the aesthetics of wood. Conversely, semi-transparent and clear coatings maintain
the natural appearance of wood, but the UV/visible light transparency of the coating
results in the photodegradation of wood substrate and the resultant failure of the coating
adhesion due to the photodegraded substrate [47]. The addition of photo stabilizers such

as UVA into coating reduces the discoloration of coating and wood substrate [48, 49].

51

2.3 Wood-Plastic Composites

2.3. 1 Introduction

The properties of WPCs combine those of wood ﬁber and polymer matrix. The
wood content in the WPCs, the characteristics of the wood and polymer, the interface
between the wood and polymer matrix, additives used in the WPC’s formulation, as well
as processing procedures are factors affecting end-use properties of WPCs. In addition,

moisture absorption and weathering of WPCs are important issues for WPC application.

2.3.2 Effect of wood content

Wood ﬁber content in WPCs signiﬁcantly inﬂuences both processing and end-use
PI‘Operties. In the melting state, the viscosity of composites increases with ﬁber loading
Since ﬁber hinders the movement of polymer chain [50-52]. Generally, the mechanical
Properties of composites follow the law of mixture. Due to higher strength but lower
mOdqus of thermoplastic compared to wood, generally increasing the wood content up to
40% decreases tensile and ﬂexural strength but increases tensile and ﬂexural modulus
[5 1 -53]. Moreover, because of hydrophilic nature of wood component and hydrophobic

nature of thermoplastic, moisture sorption increases with the wood component ratio in the

WPCs [54,55].

2.3.3 Effect of the characteristics of the wood fiber

The wood used in WPCs is generally in the form of ﬂour or ﬁber. The typical

length-to-diameter (LID) ratio of wood ﬂour ranges from about 3:1 to 5:1. The wood

52

ﬂour is graded by mesh size. For example, wood ﬂour with 40 and 80 mesh corresponds
to 0.35 and 0.18 mm, respectively. Wood ﬁber typically has higher L/D ratio than wood

ﬂour from 10:1 to 25:1. Fiber length varies from 1 to 10 mm.

The size and geometry of wood ﬁber have signiﬁcant effect on moisture
absorption and mechanical properties [56-5 8]. Moisture absorption increases with length-
to-diameter (L/D) ratio of wood ﬁber [56]. Longer ﬁbers may increase the probability of
voids within the composite. More voids would contain more water content when
composites are exposed to a humid environment. In contrast, large L/D ratio of ﬁber
favors the mechanical properties such as tensile and ﬂexural properties, since increasing
the ﬁber L/D ratio can increase the transfer efﬁciency of loading from polymer matrix to

ﬁber [56-58].

2.3.4 Effect of the characteristics of the polymer matrix

Characteristics of polymers, including molecular weight and its distribution,
rheology properties (MFI and melting point, etc.), mechanical properties, and degradation

temperature, inﬂuence the properties and/or manufacturing of WPCs.

Melt ﬂow index (MFI) of a polymer, an assessment of average molecular mass
and inverse measure of the melt viscosity, is critical to control the processing. Injection
molding encompasses higher MFI polymers while extrusion processes use lower MFI
polymers. The MFI of a polymer has a complicated effect on the mechanical properties of
WPCs. Because polymers with a low MFI often have a high molecular weight, WPCs
with a low MFI polymer matrix have better strength and modulus than WPCs with a high

MFI polymer matrix [59]. Conversely, at a high wood ﬁber loading, high MFI polymer

53

matrix improves the wood ﬁber distribution and wetting, thus results in better mechanical

properties [5 1].

2.3.5 Effect of surface modiﬁcation of wood and addition of coupling agents

Because of wood’s hydrophilic nature and plastic’s hydrophobic nature, WPCs
often require surface modiﬁcation of wood or the addition of coupling agents to decrease
the hydrophilicity of wood and increase the compatibility between wood ﬁber and
plastics [60-64]. Coupling agents or surface modiﬁers contain two domains. One domain
is a chemical function group, which can form covalent bonds by reacting with hydroxyl
group on the wood; the other domain is a long chain, which can form entanglements with
polymer matrix. In this way, the hydrophilicity of wood is decreased, as well as the
dispersion of wood in the polymer matrix and the adhesion between wood and polymer
are improved. As a result, moisture sorption is reduced and mechanical properties
sensitive to the interfacial adhesion such as ﬂexural and tensile strength are enhanced.
For example, Silane or stearic acid surface modiﬁed natural ﬁber-ﬁlled PP (50% ﬁber
loading) decreased approximately 40% moisture absorption compared to unmodiﬁed
ﬁber -ﬁlled PP [60, 61]. Addition of maleated PP signiﬁcantly increased over 20% the
ﬂexural and tensile strength of natural ﬁber ﬁlled-PP composites in the ﬁber loading

range between 20-50% [61, 62].

54

2.3.6 Effect of manufacturing methods

Manufacturing process can greatly inﬂuence surface quality. Study showed that
while injection molding led to WPCs with a thicker polymer-rich surface layer, the
surface of extruded WPCs, on the other hand, contained many wood particles not
encapsulated by polymer [65]. As a result, the water resistance of injection molded WPCs

was greater than that of extruded WPCs [65].

2.3.7 Moisture absorption of WPCs

Moisture absorption of WPCs would inﬂuence their dimension, mechanical
properties, and appearance. Moisture absorption reduces the adhesion between the ﬁller
and polymer matrix, a change that results in the loss of mechanical properties [54]. A
typical HDPE-based WPCs formulation containing 40% wood ﬂour exposed to 100%
relative humidity resulted in loss of 15% in tensile and 30% in ﬂexural properties [54].
Also, upon moisture absorption WPCs swell and thus alter in dimension. Study showed
that after water immersion, HDPE-based and PP-based WPCs experienced approximately
5% and 10% thickness swelling respectively [66]. Furthermore, in a high humid
environment, water absorption could initiate ﬁmgi growth, which causes the discoloration
[67]. Therefore, it is necessary to reduce or delay the moisture uptake of WPCs to

prolong the long-term performance of the products.

55

2.3.8 Photodegradation/weathering of WPCs

Because both wood and plastic suffer photodegradation, WPCs also undergo
photodegradation. The presence of wood accelerates the photodegradation of a polymer
matrix since wood is a good light absorber [68, 69]. Photodegradation of WPCs produces
changes in color, surface composition, and mechanical properties. Stark and Matuana
reported that after exposure to accelerated weathering, HDPE-based WPCs experienced
color fading, surface oxidation, and modulus loss [70-72]. Fabiyi et a1. investigated the
effects of natural and accelerated weathering on the appearance and chemical changes of
HDPE and PP-based WPCs [73]. They observed that the degree of color change/lightness,
carbonyl concentrations and lignin content increased with exposure time. PP-based
WPCs underwent quicker photodegradation in terms of lightness, carbonyl concentrations
and wood loss. The study established strong relationships between lightness and lignin
degradation, lightness and carboxylic acid concentration, as well as lightness and

esteriﬁcation.

The presence of moisture/water in the photodegradation accelerates the
weathering process. Water functions as a catalyst of an oxidative degradation. Moisture
absorption swells the wood ﬁber, facilitating light penetration. Consequentially, reports

noted increased lightening and ﬂexural properties loss in WPCs [74].

56

2.3.9 Protection against photodegradation

Current approaches to improve the UV resistance of WPCs focus on the bulk of
WPCs, i.e., incorporation of photostabilizers into the entire product [74-76]. Stark and
Matuana investigated the performance of a low molecular weight diester HALS, a high
molecular weight secondary amine triazine HALS, a benzotriazole UVA, and a colorant
as photostabilizers of HDPE-based WPCs containing 50% wood ﬂour [75]. The results
indicated that both UVA and colorant were more effective photostabilizers than HALS
with respect to preventing discoloration and ﬂexural properties loss. The low efﬁciency
of HALS was attributed to the possible reaction between the basic HALS and acidic
wood ﬂour. Muasher and Sain compared the performance of diester HALS with different
molecular weight, tertiary and secondary amine trizaine HALS as photostabilizers of
HDPE-based WPCs containing 40% wood ﬂour. Their ﬁndings position high molecular
weight diester HALS as most effective in terms of controlling long term fading and

yellowing distortions [76].

57

2.4 Testing and property evaluation

2.4.1 Weathering testing

Natural weathering is a combination of effects of sunlight (UV, visible, and IR),
water (dew, rain, snow, and humidity), temperature, oxygen, and biological
contamination or competition [30]. For WPCs, the damaging effect of sunlight is most
important. The spectral power distribution of noon midsummer sunlight (measured in
Cleveland Ohio, 1986) is shown in Figure 2.16. The duration and intensity of sunshine,
precipitation, and temperature and its ﬂuctuation are location and season dependent.
Because natural weathering is very time consuming, accelerated artiﬁcial weathering,
which reduces exposure time by increasing the effects of radiation and water, is often
used to test the durability of materials. Accelerated artiﬁcial weathering is made up of
cycles, which simulate damaging effects of sunlight, rain, and dew by artiﬁcial light
radiation, water spray, and condensation. Xenon are, UVA-340, and UVB-313 lamps are
primary radiation sources used in the accelerated artiﬁcial weathering. While Xenon arc
lamps mimic both UV and VIS parts of the solar spectra, UVA-340 and UVB-313 lamps
only simulate UV-A and UV-B regions of the sunlight [77]. Since photodegradation of
WPCs is mainly attributed to the UV-A region of the sunlight, UVA-340 lamps often

serve as a light source.

58

2.0

1.5

2

Irradiance (w/m /nm)

1.0

0.5

0.0

Figure 2.16.

 

 

, uv Region I Visible Light
I l j
- l I '
19; 09: it
I l g ,
_ I I
I I
I I
q I 0
.. l 3 :
l :
I I :
250 350 450 550 650

Wavelength (nanometers)

The sunlight spectrum [77].

59

 

2.4.2 X-ray photoelectron spectroscopy

X-ray photoelectron spectroscopy (XPS), a quantitative spectroscopic technique,
measures the elemental composition, empirical formula, chemical state and electronic
state of the elements that exist at the surface of the material. XPS spectra is obtained by
irradiating a material with a beam of aluminum or magnesium X-rays and measuring the
kinetic energy (KB) and number of electrons that escape from the top 1 to 10 nm of the

material.

A plot of the number of electrons detected (Y-axis) versus the binding energy of
the electrons detected (X-axis) charts a typical XPS spectrum. The already established
binding energies of the electron orbitals in atoms allows for identifying the atomic
composition of the sample surface. Measurement of the relative areas of the
photoelectron peaks allows for the quantitative determination of the sample’s

composition.

Each element produces a characteristic set of XPS peaks at characteristic binding

energy values that can be calculated by the following equation [78].
Ebinding = Ephoton ' Ekinetic — (1) (Equation 2-3)
where Ebinding represents the energy of the electron emitted from one electron

conﬁguration within the atom, Ephoton the energy of the X-ray photons, Ekinetic the

kinetic energy of the emitted electron as measured by the instrument and (I) the work
function of the spectrometer. Thus, a survey spectrum can be generated and sample’s
composition can be determined. A survey spectrum of wood-plastic composites is shown

in Figure 2.17.

60

 

C/S

 

 

 

 

x10
8 l 1 I r I r 1 I t 1 I
m
p—
7 . U -
6 . . -
5 . m
l '— .1
O
4 d
l'
a e ‘
i3
3 t s. hlg O ~
...e’ '~=‘ l ;
I 1, V4” 4'1 I
2 ”a.“ MVM l Wt d
10*”, pig” 1!:
1 . I]: r
: U)
1i 8
1
0 l J A r 1 J. L r 1 ”\rw‘Ifﬁw‘w IIIMUN‘; X‘

 

11001000 900 800 700 600 500 400 300 200 100 0

Binding Energy (ev)

Figure 2.17. A survey spectrum of HDPE-based WPCs (WPCs formulation: 50%

maple ﬂour, 44% HDPE, and 6% TPW104).

61

These characteristic peaks correspond to the electron conﬁguration within the atoms, e. g.,
Is, 25, 2p, 35, etc. The number of detected electrons in each of the characteristic peaks

directly relates to the amount of the element within the indicated area [78].

Chemical bonding frequently affects the binding energy of the electron orbital and
results in chemical shift in the kinetic energy of the photoelectron. These binding energy

shifts provide information of a chemical nature (such as atomic oxidation state) from the

sample surface. A high resolution spectrum of C15 region of WPCs is shown in Figure

2.18. C15 peak consists of four subpeaks, corresponding to carbon in different oxidation

states. C1 class of carbon atoms correspond to carbon atoms bonded only by carbon or
hydrogen bonds (C-C and C-H); C2 class of carbon atoms are carbon atoms bonded to a
single noncarbonyl oxygen (C-OH); C3 class of carbon atoms are carbon atoms bonded
to a carbonyl or two noncarbonyl oxygen (C=O or O-C-O); and C4 class of carbon atoms
are carbon atoms bonded to a carbonyl and a noncarbonyl oxygen (O-C=O) [68, 70, 72,

79].

62

 

mJ.\., V

 

I
1

16000

14000

I

12000

I

10000

I

8000

I

 

C/S

I

6000

4000

I

2000

I

 

 

l

300 298 296 294 292 290 288 286 284 282 280

 

 

Binding Energy (ev)

Figure 2.18. A high resolution spectrum of HDPE-based WPCs in C IS region (WPCs

formulation: 50% maple ﬂour, 44% HDPE, and 6% TPW104).

63

mt.
fit

is It

I" u
1113‘.

11101

2.4.3 Fourier transform infrared spectroscopy

Fourier Transform Infrared Spectroscopy (F TIR) is the powerful tool for
identifying the functional groups. This technique works on the fact that chemical bonds
vibrate at characteristic frequencies. When a material is irradiated with infrared radiation,
absorbed IR radiation usually excites molecules into a higher vibrational state. The
wavelength of light absorbed by a particular molecule is a function of the energy
difference between the rest and excited vibrational states. The wavelengths that are
absorbed by the sample are characteristic of its molecular structure [80]. This technique
measures the absorption of infrared radiation by the sample material versus wavelength.
The specimen's transmittance and reﬂectance of the infrared rays at different frequencies

is translated into an IR absorption plot consisting of reverse peaks.

The photodegradation of the WPCs results in decay of benzyl ring in the lignin,
loss of cellulose component, and increase in carbonyl groups, all of which can be

monitored by FTIR.

64

2.4.4 Color measurement

CIE L*a*b* (CIELAB) is color space, which describes all the colors visible to the
human eye (Figure 2.19). Color space has three coordinates, L*, a*, and b*. L“ represents
the lightness of the color (L* = 0 indicates black and L* = 100 indicates white); a*
(negative values indicate green while positive values indicate red) and b* (negative

values indicate blue and positive values indicate yellow) are the chromaticity coordinates.

The difference between colors is determined by Equation 2.4. Normally, AE of

1.0 is the smallest color difference noticeable for the human eye [81].

 

2 2

AB = \/ AL *2 +Aa * +Ab * (Equation 2.4)

65

Red

 

Yellow Black

Figure 2.19. CIE L*a*b* (CIELAB) color space [82].

66

:9

 

be
[in
at

it:

be;

511]

it

Spt

H:

the

def].

\

2.4.5 Flexural properties

Flexural properties are the ability of the material to withstand bending force. The
ﬂexural stress is a combination of tensile and compressive stresses. When composites are
bent, one side suffers maximum compressive stress, and the other side suffers maximum
tensile stress. Since one of the major applications of WPCs is decking, ﬂexural properties
are of a prime concern. Flexural strength, also known as modulus of rupture (MOR) or
bending strength, represents the highest stress experienced within the material at its
moment of rupture. The ﬂexural modulus, also called modulus of elasticity (MOE) or
bending stiffness, is the ratio of stress to strain in ﬂexural deformation. A three point
ﬂexural test is often used to determine the ﬂexural properties. The specimen lies on a
support span and the load is applied to the center by the loading nose producing three
point bending at a speciﬁed rate. The parameters for this test are the support span and the
speed of the loading, and the maximum deﬂection for the test. The selection of these
parameters is based on the test specimen thickness and is deﬁned by ASTM standard.

Flexural strength and modulus are determined by the following equations [4].

3PL

 

F lexural strength = —2 (Equation 2.5)
2bd
L2
F lexural modulus = (Equation 2.6)
2
4bd y

where P is applied load, L is support span, b is sample width, d is sample depth, and y is

deﬂection at load point.

67

2.5 Reference

10.

11.

12.

13.

14.

15.

16.

CM. Clemons, Wood-plastic composites in the United States: the interfacing of
two industries, Forest Products Journal, 52(6): 10-18 (2002).

AK. Bledzki, J. Gassan, and S. Theis, Wood-ﬁlled thermoplastic composites,
Mechanics of Composite Materials, 34(6): 563-568 (1998).

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

AA. Klyosov, Wood-plastic composites, John Wiley & Sons, Hoboken, (2007).
RD. Ebewele, Polymer science and technology, CRC Press, London, (2000).

AJ. Peacock, Handbook of Polyethylene: Structures, Properties, and Applications,
CRC Press, London, (2000).

http://en.wikipedia.org/wiki/Polyethylene

H.J. Giles Jr., J.R. Wagner Jr., and EM. Mount III, Extrusion: The Deﬁnitive
Processing Guide and Handbook, William Andrew Publishing, New York, (2004).

W.V. Titow, PVC Technology, Elsevier Applied Science Publishers, New York
(1984)

G. Butters, Particulate Nature of PVC, Applied Science Publishers LTD, London,
(1982)

CH. Chen, C.F. Mao, and Y.W. Lo, Study of fusion percolation thresholds of rigid
PVC compounds, Journal of Applied Polymer Science, 81(12): 3022-3029 (2001).

L.M. Matuana and J .W. Kim, Fusion characteristics of wood-ﬂour ﬁlled rigid PVC
by torque rheometry, Journal of Vinyl & Additive Technology, 13(1): 7-13 (2007).

J.F. Rabek, Polymer Photodegradation: Mechanisms and Experimental Methods,
Chapman & Hall USA, New York (1995).

NS. Allen, Degradation and Stabilisation of Polyoleﬁnes, Applied Science
Publishers, New York (1983).

P. Gijsman, G. Meijers, and G. Vitarelli, Comparison of the UV-degradation
chemistry of polypropylene, polyethylene, polyamide 6 and polybutylene
terephthalate, Polymer Degradation and Stability, 65(3): 433-441 (2006).

S. H. Hamid, Handbook of polymer degradation, Marcel Dekker, New York,
(2000)

68

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.
29.

30.

31.

J.R. White and AV. Shyichuk, Macromolecular scission and crosslinking rate
changes during polyoleﬁn photo-oxidation, Polymer Degradation and Stability,
92(7): 1161-1168 (2007).

LC. Mendes, E.S. Ruﬁno, O.C. Paula and AC. Torres. Mechanical, thermal and
microstructure evaluation of HDPE after weathering in Rio de Janeiro City,
Polymer Degradation and Stability, 79(3): 371-383 (2003).

J. Attwood, M. Philip, A. Hulme, G. Williams, and P. Shipton, The effects of
ageing by ultraviolet degradation of recycled polyoleﬁn blends, Polymer
Degradation and Stability, 91(12): 3407-3415 (2006). '

I. Yakimets, D. Lai, and M. Guigon, Effect of photo-oxidation cracks on behaviour
of thick polypropylene samples, Polymer Degradation and Stability, 86(1): 59-67
(2004)

CA. Harper, E.M. Petrie, Plastics materials and processes: a concise encyclopedia,
John Wiley and Sons, Hoboken, (2003).

F. Gugumus, Possibilities and limits of synergism with light stabilizers in
polyoleﬁns 2. UV absorbers in polyoleﬁns, Polymer Degradation and Stability,
75(5): 309-320 (2002).

C. Schaller, D. Rogez, A. Braig, Hydroxyphenyl-s-triazines: advanced multi-
purpose UV-absorbers for coatings, Journal of Coatings Technology and Research,
5(1), 25-31(2008).

F. Gugumus, Possibilities and limits of synergism with light stabilizers in
polyoleﬁns 1. HALS in polyoleﬁns, Polymer Degradation and Stability, 75(2): 295-
308(2002)

C. Schaller, D. Rogez, A. Braig, Hindered amine light stabilizers in pigmented
coatings, Journal of Coatings Technology and Research, 6(1), 81-88 (2008).

Forest Products Laboratory, Wood handbook—wood as an engineering material,
US. Department of Agriculture, Madison, (1999).

RM. Rowell, Handbook of wood chemistry and wood composites, CRC Press, New
York, (2005).

http://en.wikipedia.org/wiki/Lignin

M.D. Donohue and G.L. Aranovich. Classiﬁcation of Gibbs adsorption isotherms,
Advances in Colloid and Interface Science, 76-77: 137-152 (1998).

WC. Feist and D.N.-S. Hon, Chemistry of weathering and protection, US.
Department of Agriculture, Madison, (1984).

D.N.-S. Hon and N. Shiraishi, Wood and cellulosic chemistry, Marcel Dekker INC,
New York, (2000).

69

32.
33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

http://en.wikipedia.org/wiki/Photochemistry

B. George, E. Suttie, A. Merlin, and X. Deglise, Photodegradation and
photostabilisation of wood—the state of the art, Polymer Degradation and Stability,
88(2): 268-274 (2005).

U. Muller, M. Ratzsch, M. Schwanninger, M. Steiner, and H. Zobl, Yellowing and
IR-changes of spruce wood as result of UV-irradiation, Journal of Photochemistry
and Photobiology B Biology, 69(2): 97-105 (2003).

L. Tolvaj and O. Faix. Artiﬁcial aging of wood monitored by DRIFT spectroscopy
and CIE L*a*b* color measurements 1: effect of UV light. Holzforschung, 49(5):
397-404 (1995).

K.K. Pandey, Study of the effect of photo-irradiation on the surface chemistry of
wood, Polymer Degradation and Stability, 90(1):9-20 (2005).

X. Colom, F. Carrillo, F. Nogues, and P. Garriga, Structural analysis of
photodegraded wood by means of FTIR spectroscopy, Polymer Degradation and
Stability, 80(3):543-549 (2003).

K.K. Pandey, A note on the inﬂuence of extractives on the photo-discoloration and
photo-degradation of wood, Polymer Degradation and Stability, 87(2):375-379
(2005)

P. Nzokou and D. P. Kamdem, Inﬂuence of wood extractives on the photo-
discoloration of wood surfaces exposed to artiﬁcial weathering, Color Research &
Application, 31(5): 425-434 (2006).

Y. Kataoka and M. Kiguchi, Depth proﬁling of photo-induced degradation in wood
by FT-IR, Journal of Wood Science, 47(4): 325-327 (2001).

K. Mitsui and S. Tsuchikawa, Low atmospheric temperature dependence on
degradation of wood, Journal of Photochemistry and Photobiology B Biology,
81(1): 84-88 (2005).

D.N. -S. Hon, S.T. Chang, and WC. Feist, Participation of singlet oxygen in the
photodegradation of wood surfaces, Wood Science and Technology, 16(3) : 193-201
(1982)

MD. Meijer, Review on the durability of exterior wood coatings with reduced
VOC-content, Progress in Organic Coatings, 43(4): 217-225 (2001).

Y. Xie, A. Krause, C. Mai, H. Militz, K. Richter, K. Urban, and RD. Evans,
Weathering of wood modiﬁed with the N-methylol compound 1,3-dimethylol-4,5-
dihydroxyethyleneurea, Polymer Degradation and Stability, 89(1): 189-199 (2005).

M. Kiguchi and RD. Evans, Photostabilisation of wood surfaces using a grafted
benzophenone UV absorber, Polymer Degradation and Stability, 61(1) 33—45
(1998)

70

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

RD. Evans, N.L. Owen, S. Schmid, and RD. Webster. Weathering and
photostability of benzoylated wood, Polymer Degradation and Stability, 76(1) 291—
303 (2002).

ST. Chang and PL. Chou, Photodiscoloration of UV-curable acrylic coatings and
the underlying wood, Polymer Degradation and Stability, 63(3): 435-439 (1999).

ST. Chang and PL. Chou, Photodiscoloration inhibition of wood coated with UV-
curable acrylic clear coatings and its elucidation, Polymer Degradation and
Stability, 69(3): 355-360 (2000).

PL. Chou, H.T. Chang, T.F. Yeh, and ST. Chang, Characterizing the conservation
effect of clear coatings on photodegradation of wood, Bioresource Technology,
99(5): 1073-1079 (2008).

P.Y. Kuo, S.Y. Wang, J.H. Chen, H.C. Hsueh, M.J. Tsai, Effects of material
compositions on the mechanical properties of wood-plastic composites
manufactured by injection molding, Materials and Design, 30(9): 3489-3496
(2009)

P. W. Balasuriya, L. Ye, and Y.W. Mai, Mechanical properties of wood ﬂake—
polyethylene composites. Part 1: effects of processing methods and matrix melt ﬂow
behaviour, Composites Part A: Applied Science and Manufacturing, 32(5): 619-629
(2001).

S. Narongrit, C. Kantima, P. Chakarin, and T Sirinthom, Effect of wood sawdust
content on rheological and structural changes, and thermo-mechanical properties of
PVC/sawdust composites, Polymer International, 52(12): 1847-1855 (2003).

A. J. Nufiez, P. C. Sturm, J.M. Kenny , M. I. Aranguren , N. E. Marcovich , M. M.
Reboredo, Mechanical characterization of polypropylene-wood ﬂour composites,
Journal of Applied Polymer Science, 88(6): 1420-1428 (2003).

N.M. Stark, Inﬂuence of moisture absorption on mechanical properties of wood
ﬂour-polypropylene composites, Journal of Thermoplastic Composite Materials,
14(5): 421-432 (2001).

NE. Marcovich, M.M. Reboredo, and MI. Aranguren, Moisture diffusion in
polyester-wood ﬂour composites, Polymer, 40(26): 7313-7320 (1999).

S. Migneault, A. Kouba, F. Erchiqui, A. Chaala, K. Englund, and MP. Wolcott,
Effects of processing method and ﬁber size on the structure and properties of
wood—plastic composites, Composites Part A: Applied Science and Manufacturing,
40(1): 80-85 (2009).

H.C. Chen, T.Y. Chen, and CH. Hsu, Effects of wood particle size and mixing
ratios of HDPE on the properties of the composites, Holz als Roh- und Werkstoff,
64(4): 172-177 (2006).

71

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

M. C. N. Yemele, A. Koubaa, A. Cloutiera, P. Soulounganga, and M. Wolcott,
Effect of bark ﬁber content and size on the mechanical properties of bark/HDPE
composites, Composites Part A: Applied Science and Manufacturing, 41(1): 131-
137(2010)

H. Huang and J. Zhang, Effects of ﬁller—ﬁller and polymer—ﬁller interactions on
rheological and mechanical properties of HDPE-wood composites, Journal of
Applied Polymer Science, 111(6): 2806-2812 (2009).

S. Lee, S. Chun, G. Doh, and I. Kang, Inﬂuence of chemical modiﬁcation and ﬁller
loading on fundamental properties of bamboo ﬁbers reinforced polypropylene
composites, Journal of Composite Materials, 43(15): 1639-1657 (2009).

L. Danyadi, J. Moczo, and B. Pukanszky, Effect of various surface modiﬁcations of
wood ﬂour on the properties of PP/wood composites, Composites Part A: Applied
Science and Manufacturing, 41(2): 199-206 (2010).

T.J. Keener, R.K. Stuart, and T.K. Brown, Maleated coupling agents for natural
ﬁbre composites, Composites Part A: Applied Science and Manufacturing, 35(3):
357-362 (2004).

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

J.Y. Kim, J.H. Peck, S.H. Hwang, J. Hong, S. C. Hong, W. Huh, S.W.Lee,
Preparation and mechanical poperties of Poly(vinyl chloride)/bamboo ﬂour

composites with a novel block copolymer as a coupling agent, Journal of Applied
Polymer Science, 108(2): 2654—2659 (2008).

N.M. Stark, L.M. Matuana, and CM. Clemons, Effect of processing method on
surface and weathering characteristics of wood—ﬂour/I-IDPE composites, Journal of
Applied Polymer Science, 93(3): 1021-1030 (2004).

KB. Adhikary, S. Pang, and MP. Staiger, Long-term moisture absorption and
thickness swelling behaviour of recycled thermoplastics reinforced with Pinus
radiata sawdust, Chemical Engineering Journal, 142(2) 190-198 (2008).

B. Dawson-Andoh and L.M. Matuana, High density polyethylene-wood ﬂour
composite lumber: efﬁcacy of two commercial biocides in the control of fungal
colonization and discoloration, Holz als Roh-und Werskstoff, 65(5): 331-335
(2007)

L.M. Matuana and DP. Kamdem, Accelerated ultraviolet weathering of
PVC/wood-ﬂour composites, Polymer Engineering and Science, 42(8): 1657-1666
(2002)

L.M. Matuana, D.P. Kamdem, and J. Zhang, Photoaging and Stabilization of rigid
PVC/wood-ﬁber composites, Journal of Applied Polymer Science, 80(11): 1943-
1950(2001)

72

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.
81.

82.

NM. Stark and L.M. Matuana, Characterization of weathered wood—plastic
composite surfaces using FTIR spectroscopy, contact angle, and XPS, Polymer
Degradation and Stability, 92(10): 1883-1890 (2007).

N.M. Stark and L.M. Matuana, Surface chemistry and mechanical property changes
of wood-ﬂour/high-density-polyethylene composites after accelerated weathering,
Journal of Applied Polymer Science, 94(6): 2263-2273 (2004).

N.M. Stark and L.M. Matuana, Surface chemistry changes of weathered
HDPE/wood-ﬂour composites studied by XPS and FTIR spectroscopy, Polymer
Degradation and Stability, 86(1): 1-9 (2004).

J.S. Fabiyi, A.G. McDonald, M. P. Wolcott, and P. R. Grifﬁths, Wood plastic
composites weathering: visual appearance and chemical changes, Polymer
Degradation and Stability, 93(8): 1405-1414 (2008).

N.M. Stark and L.M. Matuana, Inﬂuence of photostabilizers on wood ﬂour-HDPE
composites exposed to xenon-arcradiation with and without water spray, Polymer
Degradation and Stability, 91(12): 3048-3056 (2006).

N.M. Stark and L.M. Matuana, Ultraviolet weathering of photostabilized wood-

ﬂour-ﬁlled high density polyethylene composites, Journal of Applied Polymer
Science, 90(10): 2609-2617 (2003).

M. Muasher and M. Sain, The efﬁcacy of photostabilizers on the color change of
wood ﬁlled plastic composites, Polymer Degradation and Stability, 91(5): 1156-
1165 (2006).

P. Brennan and C. Fedor, Sunlight, UV and accelerated weathering, Technical
Bulletin, The Q-Panel Company, (1987).

M.Stamm, Polymer surfaces and interfaces: characterization, modification and
applications, Springer, Dresden, (2008).

G. Sinn, A. Reiterer and S. E. Stanz-Tschegg, Surface analysis of different wood
species using X-ray photoelectron spectroscopy (XPS), Journal of Materials
Science, 36(19): 4673-4680 (2001).

J .E. Mark, Physical properties of polymers handbook, Springer, New York, (2006).

American Society for Testing and Materials, E308-08 Standard practice for
computing the colors of objects by using the CIE system, in Annual Book of ASTM
Standards, 06.01 , (2008).

N.M. Stark, Photodegradation and photostabilization of weathered wood ﬂour
ﬁlled polyethylene composites, PHD Dissertation, Michigan Technological
University, (2003).

73

CHAPTER 3

Coextruded PVC/Wood-Flour Composites with WPC Cap Layers

This chapter is slightly modiﬁed from a research paper that was published in the Journal
of Vinyl & Additive Technology, 14 (4): 197-203 (2008). It is co-authored by Shan Jin

and L.M. Matuana.

74

3.1 Abstract

Wood-plastic composites (WPCs) can absorb moisture in a humid environment
owing to the hydrophilic nature of the wood, thereby making the products susceptible to
microbial growth and loss of mechanical properties. In this study, rigid poly(vinyl
chloride) (PVC)/wood-ﬂour composites (core layer) were coextruded with either unﬁlled
rigid PVC (cap layer) or rigid PVC ﬁlled with a small amount (5—27.5%) of wood ﬂour
(composite cap layers) in order to decrease or delay the moisture uptake. The thickness of
the cap layer and its composition in terms of wood ﬂour content were the variables
examined during coextrusion. Surface color, moisture absorption, and ﬂexural properties
of both coextruded and noncoextruded (control) composite samples were characterized.
The experimental results indicated that both unﬁlled PVC and composite cap layers can
be encapsulated over rigid PVC/wood-ﬂour composites in a coextrusion process. The
moisture uptake rate was lower when a cap layer was applied in the composites, and the
extent of the decrease was a strong function of the amount of wood ﬂour in the cap layer
but insensitive to cap layer thickness. Overall, coextruding PVC surface-rich cap layers
on WPCs signiﬁcantly increased the ﬂexural strength but decreased the ﬂexural modulus
as compared with those of control samples. The changes in bending properties were

sensitive to both cap layer thickness and wood ﬂour content.

75

3.2 Introduction

Wood-plastic composites (WPCs) have experienced signiﬁcant market expansion
in recent years as a replacement for solid wood, mainly in outdoor applications such as
railings, decking, fences, window and door frames etc. These products are marketed as
low-maintenance and weather-resistant compared with solid wood, because the wood can
be encapsulated in a hydrophobic polymer matrix [1, 2]. However, as WPCs are exposed
to an outdoor environment of moisture and ultraviolet light in sunlight, the expectation
that the products would maintain their new appearance without maintenance has not been

realized [1-3].

Indeed, studies have shown that WPCs can absorb moisture in a humid
environment because of the highly hydrophilic nature of wood ﬂours in the composites,
especially at high loading levels, thereby making the products more prone to microbial
growth and biological decay, and thus affecting their appearance [4, 5]. Moisture
absorption also reduces the adhesion between the ﬁller and polymer matrix, a change that
results in the loss of mechanical properties [3]. Therefore, it is necessary to reduce or

delay the moisture uptake of WPCs to prolong the long term performance of the products.

Many researchers have investigated the moisture/(water absorption) behaviors of
natural ﬁber—plastic composites. Marcovich et a1. investigated moisture diffusion in
polyester-(wood ﬂour) composites and found that the equilibrium moisture content of the
composites was dependent on the size and shape of the wood ﬂour [6]. Stark studied the
inﬂuence of moisture absorption on the mechanical properties of (wood ﬂour)-
polypropylene composites [3]. She reported that incomplete encapsulation of wood ﬂour

by the polypropylene matrix occurs in both 20% and 40% wood ﬂour composites. In

76

addition, she demonstrated that more moisture was absorbed by the composites with
higher wood ﬂour loading, a situation that resulted in a signiﬁcant decrease of both
ﬂexural and tensile properties [3]. Similar results were reported by Bledzki and Faruk,
who found that the impact properties of (wood ﬁber)—polypropylene composites were
affected by moisture content [7]. Hygrothermal aging behavior of injection-molded (rice
husk)-ﬁlled polypropylene composites also has been examined by immersing the
specimens in distilled water at 30°C and 90°C [8]. It was found that both the tensile
strength and the modulus deteriorated after immersion in water, the extent of the
deterioration being dependent on the immersion temperature. This reduction in tensile

properties was attributed to the interfacial degradation in the composites [8].

Although several authors have investigated the moisture/(water absorption)
behaviors of (natural ﬁber)-plastic composites, only a few provided the methods to
alleviate this problem. The use of additives, such as maleated polyoleﬁns and other
compatibilizers, encapsulation of exposed wood during compounding, or the use of
protective coating have been reported as methods to improve the moisture uptake of
WPCs [9]. These suggested approaches are employed by current WPC manufacturers
with limited success. Stark and Matuana have recently demonstrated that covering the
WPCs with hydrophobic layers decreases the moisture absorption of WPCs. Their results
showed that a cap layer of polyoleﬁn coextruded over (wood ﬂour)-polyoleﬁn
composites provides improved moisture resistance compared with that of the control

WPC [10].

Coextrusion technology can produce a multilayered product with different

properties at outer and inner layers, thus offering different properties between surface and

77

bulk [1 1-13]. The beneﬁt of coextrusion is that each layer of coextruded material offers a
particular property, including impermeability or resistance to moisture uptake and
improved mechanical properties, such as stiffness, depending on the compositions of the
outer and inner layers. For example, when composites are bent, one side suffers
maximum compressive stress, and the other side suffers maximum tensile stress [13].
Therefore, the composition of the outer layer will inﬂuence not only the moisture uptake,

but also the ﬂexural properties of the overall composite.

The main goal of this study was to coextrude rigid PVC/wood-ﬂour composites
with various cap layers in order to decrease or delay the moisture uptake of the
composites and thus increase the service life of the WPCs. Particular emphasis was
placed on examining the effects of cap layer composition (unﬁlled PVC vs. WPC) on

moisture uptake and ﬂexural properties.

78

3.3 Experimental

3.3.1 Materials

The PVC matrix and other additives used in the manufacture of the coextruded
composites are listed in Table 3.1. Rutile titanium dioxide supplied by Kemira was used
as a pigment in the cap layer formulations to differentiate the cap layer from the core. A
0.425 mm (40 mesh) maple wood ﬂour was utilized as ﬁller. The wood ﬂour was oven-
dried at 105°C for 48 h to remove moisture before processing. Formulations of core layer
(rigid PVC/wood-ﬂour composites) and cap layers [unﬁlled rigid PVC and rigid PVC

ﬁlled with a small amount of wood ﬂour (5—27.5 wt%)] are also listed in Table 3.1.

3.3.2 Compounding and coextrusion of composites

A 20-L high-intensity mixer [Papenmeier, Type TGAHK20] was used for dry

blending of the TiOg-pigmented pigmented rigid PVC as well as unpigmented PVC, dried

maple wood ﬂour, and other additives at the concentration levels listed in Table 3.1. Six
different formulations were compounded in order to produce control and coextruded

samples with different types of cap layers. These formulations consisted of:

o 100 phr of nonpigmented rigid PVC and 100 phr of wood ﬂour as control (no cap
layer) and core layer,

0 unﬁlled rigid PVC pigmented with 5 phr of Ti02 for cap layer, and

o pigmented rigid PVC ﬁlled with four different amounts of wood ﬂour (5, 12.5, 20,

and 27.5 wt%) for cap layers.

79

Table 3.1. Formulations used in core and cap layers of coextruded rigid PVC/wood-

 

 

 

 

ﬂour composites.
Content (phr)
Ingredients Core Layer Cap layers
and control
PVC (K-value = 66.4) (Shintech) 100 100
Tin stabilizer (PlastiStab 2808-Halstab) 2 2
Calcium stearate (Synpro) 1.5 1.5
Parafﬁn wax (Gulf Wax) 2 2
Processing aid (Paraloid K-120)l 2 2
Processing aid (Paraloid K-175)l 2 2
Impact modiﬁer (K-334)' 10 10
40 mesh maple (American Wood Fibers)

Titanium dioxide (Kemira 460)

100 (0-27.5 wt%)

 

1Supplied by Rohm & Haas Co.

80

Each composition (Table 3.1) was dry blended for 10 min, and the shear heat
generated during compounding increased the temperature of the blend from 24°C to 35°C.
The compounded materials were discharged to a container at room temperature. The
coextrusion system consisted of a coextrusion die connecting a twin-screw extruder for
the core layer and a single-screw extruder for the cap layer. Core layer was formed by
using a 32-mm conical counter rotating twin screw extruder (C.W. Brabender
Instruments, South Hackensack, NJ) with a length-to-diameter ratio of 13:1. Cap layers
were produced by using a 19-mm single-screw extruder (C.W. Brabender Instruments,
South Hackensack, NJ) with a length-to-diameter ratio of 30:1. The die used in our
experiments was speciﬁcally designed to ensure that the cap layers covered the entire
core layer of the WPCs, i.e., to ensure complete encapsulation of the core layer. The

schematic of the coextrusion system used in this study is shown in Figure 3.1.

81

  
 
    
   
 

   
  

I'\\‘|'
AL

Twin screw extruder

/ (core layer)

 
         
         
       
 
  

/

Single screw extruder
(cap layer)

     
 

Die

 

Coextruded sample
(Fully capped)

Core layer
Cap layer

Figure 3.1. Schematic of the coextrusion system used in this study.

82

Core layers consisted of unpigmented rigid PVC/wood-ﬂour composites made
with 100 phr of PVC and 100 phr of wood ﬂour (Table 3.1). It was observed in our
previous work that the surface quality of rigid PVC/wood-ﬂour composites was largely
dependent on the processing conditions, particularly the barrel temperature proﬁle and
screw rotational speed of the extruder as well as on the amount of wood ﬂour used [1].
For example, samples of excellent quality (i.e., smooth without torn edges) were
produced for rigid PVC ﬁlled with maple ﬂour up to 100 phr by using a decreased
temperature proﬁle from the hopper to the die. Consequently, the processing temperature
proﬁle of the twin-screw extruder for the core layer was set at 190/ 180/ 175/ 170°C from
hopper to coextrusion die. The rotational speed of the screws was maintained at 30 rpm
during processing. These processing conditions for the core layer were used in the

manufacture of all coextruded composites.

By contrast, various formulations and processing conditions were utilized during
the coextrusion process for the cap layers. When unﬁlled rigid PVC was used as the cap
layer, the processing temperature proﬁle of the single screw extruder was set at
155/ 165/165/ 170°C from hopper to coextrusion die. However, three different rotational
screw speeds (5, 10, and 15 rpm) were employed to vary the thickness of the cap layers.
On the other hand, the processing temperature proﬁle of the single-screw extruder for
rigid PVC/wood-ﬂour composites cap layers (or composite cap layers) was set at
185/ 180/ 175/ 170°C from hopper to coextrusion die, regardless of the formulations.
Similarly, the rotational screw speed was maintained constant at 15 rpm throughout the
experiments when cap layers of composites were coextruded over WPCs. Nevertheless,

the concentrations of wood ﬂour in the cap layers varied from 5 to 27.5% based on the

83

total weight of the cap layers of the composite. The control samples (rigid PVC/wood-
ﬂour composites without cap layer) were also produced by using the coextrusion die. The

processing conditions were same as those utilized for core layers.

3.3.3 Property evaluation
Cap layer thickness evaluation

Pictures of the cross sections of coextruded samples were taken by using a digital
camera, and samples were visually examined to evaluate the thickness of the cap layer
(Table 3.2). Figure 3.2 shows the cross section of unﬁlled rigid PVC coextruded on

WPCs.

Color measurements

Color measurements on the surfaces of control and coextruded samples were
made according to the procedure outlined in ASTM D2244. A Minolta CR-42O Chroma
Meter (Minolta Corp., Ramsey, NJ) was used to measure the color in L*a*b* coordinates
at three locations on each sample by using the Commission International d’Eclairage
(CIELAB 1976 color space) three-dimensional color space system. In this system, L*axis
(+L* for light, -L* for dark) represents the lightness, and a* (+a* for red, -a* for green),

and b* (+b* for yellow, -b* for blue) are the chromaticity coordinates.

84

.5852? of E 83098 mo Amman mde n vm _ w beachioﬁam Sam team—=28 mm?» EuEOQ 2:568 ushm

.moEEmm =a 8m 58 $8 ES _.o 83 3053:: Br?— Qmo com :owaQv c.8983 2;.

 

 

 

- - 86 35.. mg a 3 Age 3 SE 98 vi
2 a: one :3- 3a a we EB o: 53 a8 oi
S 8; 8+ 3. _- 25 E go A83 a 53 98 vi

mwoomm mammmooem
3 N3 N3 one ma 2 2 Son 825 exam
3 a: 31m “2 mi 2 2 so: 803 £92
”a 2.». NS 2: n? 3 3 Bow Boa $8
2 - a.” own NE 3 A: Bow 895 :98
3 3m ~52 a: new - - 29:58 082
3.x. we: ~52 N g 22 .3 ...a L =5 as $5 98 a
.mmoqxomﬁ Hot?— Eoucoo Soc coo?
E80833 coup—om ESQEEE 830 98 owSo><

 

@2953 .8280 98 3:89:00 BocéootsCx/m Em: 39558

mo .539on 03368 was .830 ooﬂbﬁ £3535 m: :0 Hot?— 93 mo 3on mammmoooa was coEmOQEoo 05 mo muoobm

.m.m 2an

85

 

Figure 3.2. Cross section of unﬁlled rigid PVC cap layer coextruded on WPCs.
The arrows show the areas of poor interfacial adhesion between cap

and core layers.

86

Moisture absorption tests

The moisture absorption tests were carried out by following the method described
by Kim et a1. [14]. In these tests, a saturated ammonium chloride solution was placed in a
sealed chamber to provide a constant relative humidity of 79% at 20°C, which yielded an
equilibrium moisture content of approximately 16%. The nominal width, depth and
length of the samples were 2.54 cm (1 inch), 0.95 cm (3/8 inch), and 7.62 cm (3 inch),
respectively. Both dried control and coextruded composite samples with unsealed edges
were weighed and placed in this sealed chamber. Samples were removed from the
chamber and weighed again at various intervals to record mass gained. Theopercentage of

moisture absorbed by the sample or moisture content (MC) was calculated from the

recorded mass of the sample at time t (M0 and the initial mass of the dry sample (M0) by

using the following equation:
M — M
MC = ;M—O— x100 (Equation 3.1)

The mass of specimens was measured to a precision of 0.0001 g. At least 3 replicates

were used to obtain an average value for each formulation.

Moisture uptake curves as a function of exposure time (MC versus time) were
derived from the data obtained with Equation 3.1. From these curves, a moisture uptake

rate (MUR) was deﬁned to estimate how fast the moisture absorption occurred in the

samples:

AMC
At

MUR = (Equation 3 .2)

87

where AMC is the change in moisture content (%) and At is change in time (hours).
Values for MUR (%/h) were calculated from the initial slope of a plot of MC versus time.

Only the straight-line region below the maximum moisture content was used for this

calculation.

F lexural property evaluation

Three-point ﬂexural tests were carried out in a walk-in conditioning room at (23 d:
2)°C and (50 at 4)% relative humidity on an Instron 5585H (with Bluehill 2 software)
testing machine. Samples were stored in this conditioning room for at least 48 h before
property testing. The testing was performed with a crosshead rate of 4.3 mm/min in
conformance with ASTM standard D7031-2004, the standard test method for evaluating
the mechanical and physical properties of WPC products. At least six replicates were
tested to obtain an average value for each formulation. Data were collected on modulus
of rupture (MOR or ﬂexural strength) and modulus of elasticity (MOE or ﬂexural

stiffness).

Statistical analysis

A two-way analysis of variance (ANOVA) was performed on the MOR and MOE
data to distinguish statistical differences (p < 0.05) between coextruded and control
samples. All statistical analyses were performed by using Design Expert software (v. 7)
from Stat-Ease (Minneapolis, MN). The superscript letters in Tables 3.3 and 3.4 represent

the ANOVA results. The same letter denotes that the difference between two treatments

88

is not statistically signiﬁcant. Otherwise, the difference is statistically signiﬁcant (p <

0.05).

89

3.4 Results and Discussion

3.4.1 Color measurements

As illustrated in Figure 3.1, the goal of coextrusion was to encapsulate rigid
PVC/wood-flour composites completely with both unﬁlled rigid PVC and rigid PVC
ﬁlled with a small amount of wood ﬂour (or composite cap layers). An image of the cross
section of a coextruded sample is shown in Figure 3.2, and the average thicknesses of the
cap layers are listed in Table 3.2. The thickness of the cap layer was affected by different
rotational screw speeds of the extruder feeding the die for cap layers. Increasing the
rotational screw speed tended to increase the thickness of the cap layer, mainly owing to
the increased output during coextrusion processing. However, the amount of wood ﬂour
in the cap layers appeared to have no signiﬁcant effect on their thicknesses, probably
because all coextruded samples with composite cap layers were produced at a constant

rotational screw speed.

Color measurements were performed to monitor the change in surface color of
coextruded and control samples. The surfaces of WPC coextruded with unﬁlled and
wood-ﬂour-ﬁlled PVC cap layers are shown in Figure 3.3 and Table 3.2 summarizes the

color parameters obtained on the surfaces of coextruded and control samples.

90

 

(a) (b) (c) ((1)

Figure 3.3. WPC samples coextruded with (a) unﬁlled PVC cap layer, (b) 5%
wood ﬂour ﬁlled PVC cap layer, (c) 20% wood ﬂour ﬁlled PVC cap

layer, and (d) control WPC (no cap layer).

91

Coextruded samples with composite cap layers had higher L* but lower a* and b*
values than the composites without a cap layer (control or nonencapsulated sample),
irrespective of the wood ﬂour content in the cap layers. This result was expected, because
more wood ﬂour was exposed on the surface of the control sample, which darkened its
surface because of improper encapsulation of wood by the matrix. The extent of
darkening was dependent on the amount of wood ﬂour in the cap layer, as can be seen in
Figure 3.3. Increasing the wood ﬂour content in the cap layers tended to darken the
surface of coextruded composites owing to the darker color of wood flour. As expected,
the surface of the composites became lighter when TiOz-pigmented unﬁlled PVC cap
layers were coextruded over rigid PVC/wood-ﬂour composites, regardless of the

thickness of the cap layer.

These experimental results imply that composite cap layers which contain a small
amount of wood ﬂour or a thin rigid PVC cap layer can be encapsulated over rigid

PVC/wood-ﬂour composites in a coextrusion process.

92

3.4.2 Moisture absorption

Moisture absorption curves as a function of exposure times of various cap layers
(unﬁlled PVC and composites) coextruded over rigid PVC/wood-ﬂour composites
(WPCs) are shown in Figure 3.4. The sorption parameters estimated from these curves,
i.e., maximum moisture content and the rate of moisture uptake determined from

Equation 3.2, are summarized in Table 3.2.

It should be mentioned that the samples did not reach the equilibrium moisture
content even after approximately 8124 b (338.5 days) of exposure in the chamber,
probably because of their larger dimensions. However, the absence of distinct plateau
regions in Figure 3.4 does not affect the initial slope of the curves used to calculate the
rate of moisture uptake. In addition, the linear region is not clearly seen in the moisture
absorption curves of coextruded samples with unﬁlled rigid PVC cap layers of various
thicknesses (Figure 3.4b), mainly because of the magnitude of the moisture content
absorbed by nonencapsulated WPCs (control sample). Instead, a clear linear region could

be seen for these samples by discarding the curve of the control sample in this ﬁgure.

93

Figure 3.4:

Moisture content (%)

Moisture content (%)

Moisture absorption curves of control samples and coextruded rigid
PVC/wood-ﬂour composites with various types of cap layers: (a) wood-

ﬂour (5-20 wt%) ﬁlled rigid PVC cap layers and (b) unﬁlled rigid PVC

 

   
   

 

 

 

   

 

     

 

+ Control (no cap) (a)
4 + —0 -20% wood ﬂour
- 4- - 12.5% wood ﬂour
3 A, —A— 5% woodflour ’ of
t
1
0 i ‘ ‘ A l L TL ‘ if +
0 1500 3000 4500 6000 7500 9000
Time (hours)
5
l —o—Control (no cap) (b)
4 —~ —D- -0.6 mm PVC cap
- -I- - 0.8 mm PVC cap
3 t
2 ..
1 . i" 1 a?- ‘
O 1 ‘ t , t * i ‘ 2 4r

 

 

0 1500 3000 4500 6000 7500 9000

Time (hours)

cap layers of various thicknesses.

94

Nonencapsulated WPCs (control) not only absorbed more moisture than the
coextruded counterparts after 338.5 days, but also experienced a quicker moisture uptake
than the coextruded samples, regardless of the type of cap layer. This great afﬁnity for
moisture can be attributed to the presence of hygroscopic wood on the surface of
noncoextruded WPC samples. By contrast, cap layers of both composites with a small
wood ﬂour content (Figure 3.4a) and unﬁlled rigid PVC (Figure 3.4b) signiﬁcantly
slowed the moisture uptake of WPCs, probably because of the hydrophobic layer of PVC
on the surface of the WPCs. A similar trend was reported by Stark and Matuana, who
found that the plastic-rich polymer ﬁlm that forms on the surface of coextruded WPCs

increases their moisture resistance over that of noncoextruded WPCs [10].

The extent of MUR was a strong function of the amount of wood ﬂour in the cap
layer. As shown in Figure 3.4a and Table 3.2, the MUR was lower when a cap layer was
applied in the composites. Increasing wood ﬂour content in the cap layer tended to
increase the moisture absorption rate in the WPCs. This deleterious effect can clearly be
seen in Figure 3.4a by comparing the magnitude of the slope of the curves (or the rate of
moisture uptake in Table 3.2) as well as the exposure time needed by the samples to
absorb a certain amount of moisture. As shown by the slopes in Figure 3.4a, the rate of
moisture absorption (Table 3.2) in coextruded samples decreased as the wood ﬂour
content decreased in the cap layer. It is also noted that by drawing a horizontal line
parallel to the x-axis at an arbitrary moisture content (e.g., 1% in Figure 3.4a), four
intersection points with the data can be obtained. The ﬁrst intersection point is detected at

lower exposure time (control), and the next three are seen at higher exposure time

95

(coextruded samples), thus clearly indicating the advantage of coextruding WPCs in

delaying the rate of moisture uptake.

Conversely, the extent of MUR for coextruded WPCs was not sensitive to the
thickness of the cap layer. As illustrated in Figure 3% and Table 3.2, the thickness of
unﬁlled PVC cap layers coextruded on WPCs did not affect the rate of moisture uptake of
the samples. Nevertheless, coextruded samples with a rigid PVC cap layer still absorbed
moisture, despite the presence of the hydrophobic layer of PVC on the surface of the
WPCs. As the surface of coextruded samples was covered by a PVC layer, the only
means by which they could absorb moisture was through their unprotected edges [15]. A
lack of a good interfacial adhesion between the PVC cap layer and the WPC core also
may have provided channels through which moisture diffused into the samples. As seen
at the areas close to the sample edges in Figure 3.2, the adhesion between cap and core
layers deteriorated more as the amount of wood ﬂour increased in the cap layers. As a
result, the moisture content absorbed after 8124 h by coextruded samples with only 5% of
wood ﬂour in the cap layer (Figure 3.4a) was almost 30% greater than that absorbed by

their counterparts with unﬁlled PVC cap layers.

96

3.4.3 Flexural properties

The effect of cap layer thickness on the ﬂexural properties of coextruded WPC is
summarized in Table 3.3. Signiﬁcant improvement in bending strength (MOR) occurred
in WPCs coextruded with unﬁlled rigid PVC cap layers. Increasing the cap layer
thickness did not affect the strength of coextruded WPCs. By contrast, the stiffness
(MOE) of WPCs was signiﬁcantly reduced when they were encapsulated with unﬁlled
PVC cap layers. Generally, the stiffness of coextruded WPCs decreased as the cap layer

thickness increased (Table 3.3).

The distinct trends described earlier were expected, because the mechanical
properties of coextruded samples depend on the properties of the materials in both outer
and inner layers. Generally, stronger and stiffer materials on the outer layer would
enhance the overall performance of the coextruded samples compared with that of the
product made by the inner material alone. Therefore, the increased bending strength and
decreased bending modulus induced by coextruding unﬁlled PVC cap layers over WPCs
can be attributed to greater strength but lower stiffness of unﬁlled PVC compared with

WPCs.

The effect of composite cap layer composition in terms of wood ﬂour content is
listed in Table 3.4. Composites used as cap layers contained less wood ﬂour (up to 27.5
wt%) than the one used in the core, which contained 50 wt%, thus implying that the
formulation used for the outer layer was rich in PVC. Consequently, the bending strength
of WPCs increased while their modulus decreased signiﬁcantly on encapsulating with

composite cap layers.

97

Table 3.3. Effect of the thickness of cap layer on the ﬂexural properties of

coextruded WPCs.

 

Thickness of unﬁlled PVC cap layers

 

None (Control or not coextruded)
0.6 mm (24 mil)
0.8 mm (31 mil)

1 mm (39 mil)

F lexural properties

 

MOR (MPa)

 

44.2 i 2.3A
54.9 i 1.3B
54.8 i 3.0B

53.6 i 0.7B

MOE (GPa)

 

4.52 i 0.22A
4.29 i 0.06B
3.84 i 0.24C

3.43 i 0.05D

 

The superscript letters indicate ANOVA results.

98

Table 3.4. Effects of wood ﬂour content in cap layers on the ﬂexural properties of

coextruded WPCs.

 

Wood ﬂour content (%) in cap layers

 

None (Control or not coextruded)
5% wood ﬂour
12.5% wood ﬂour
20% wood ﬂour

27.5% wood ﬂour

 

 

 

Flexural properties
MOR (MPa) MOE (GPa)
44.24 i 2.31A 4.52 : 0.22A
52.65 i 3.57B 3.29 i 0.34B
52.46 i 0.80B 3.56 50.15C
52.40 i 3.088 4.01 i 0.22D
55.23 i 2.88B 4.46 i 0.23A

 

The superscript letters indicate ANOVA results.

99

The extent of strength improvement was not a function of the wood ﬂour content
in the composite cap layer. Interestingly, the bending strengths of coextruded WPCs with
composite cap layers (Table 3.4) were similar to those produced with unﬁlled PVC cap
layers (Table 3.3). These results imply that the surface of WPCs became more
hydrophobic or PVC surface-rich after coextrusion. By contrast, the amount of wood
ﬂour in the composite cap layers affected the stiffness of coextruded WPC in two distinct
ways. The stiffness of coextruded WPCs signiﬁcantly decreased when a small amount of
wood ﬂour (5%) was used in the composite cap layer, a result which was expected owing
to the large concentration of unﬁlled PVC on the surface of the coextruded samples. An
opposite trend was observed upon increasing the wood ﬂour content, a ﬁnding which also
was expected from the rule of mixtures, i.e., the stiffness of composites increases with

increases in ﬁller content.

100

[.ij ' q

3.5 Conclusions

Rigid PVC/wood-ﬂour composites (core layer) were coextruded with either
unﬁlled PVC (cap layer) or PVC ﬁlled with a small amount (5-27.5 wt%) of wood ﬂour
(composite cap layers) in order to decrease or delay the moisture uptake of the
composites. The thickness of the cap layer and its composition in terms of wood ﬂour
content were the variables examined during the coextrusion process. The surface color,
moisture absorption, and ﬂexural properties of both coextruded and noncoextruded
(control) composite samples were characterized as functions of cap layer thickness and
amount of wood ﬂour in the cap layer. The following conclusions were drawn from the

experimental results.

Both unﬁlled PVC and composite cap layers can be encapsulated over rigid

PVC/wood-ﬂour composites in a coextrusion process.

Nonencapsulated WPCs (control) not only absorbed more moisture than the
coextruded counterparts after 338.5 days, but also experienced a quicker moisture uptake
than the coextruded samples, regardless of the type of cap layer. The MUR was lower
when a cap layer was applied in the composites, and the extent of the decrease was a
strong function of the amount of wood ﬂour in the cap layer but insensitive to the cap

layer thickness.

Overall, coextruding PVC surface-rich cap layers on WPCs has signiﬁcantly
increased their ﬂexural strength but decreased their ﬂexural modulus compared with
those of control samples. The changes in ﬂexural properties were sensitive to both the

thickness of the cap layer and its wood ﬂour content.

101

3.6 Acknowledgements

The authors are grateful for the generous donation of materials provided by the
manufacturers listed in the Experimental section and Table 3.1. The authors also are
grateful for the ﬁnancial support of USDA-CSREES Grant-Advanced Technology

Applications to Eastern Hardwood Utilization (contact grant number: 2006-34158-17581).

102

3.7 References

10.

ll.
12.

13.

J.M. Pilarski and L.M. Matuana, Durability of wood ﬂour-plastic composites
exposed to accelerated freeze-thaw cycling. Part 1. Rigid PVC matrix, Journal of
Vinyl & Additive Technology, 11(1): 1-8 (2005).

J.M. Pilarski and L.M. Matuana, Durability of wood ﬂour-plastic composites
exposed to accelerated freeze-thaw cycling. 11. High density polyethylene matrix,
Journal of Applied Polymer Science, 100(1): 35-39 (2006).

N.M. Stark, Inﬂuence of moisture absorption on mechanical properties of wood
ﬂour- polypropylene composites, Journal of Thermoplastic Composite Materials,
14(5): 421-432 (2001).

B. Dawson-Andoh, L.M. Matuana, and J. Harrison, Susceptibility of high density
Polyethylene/wood-ﬂour composite to mold discoloration, Journal of the Institute
of Wood Science, 17(2): 114-119 (2005).

B. Dawson-Andoh, L.M. Matuana, and J. Harrison, Mold susceptibility of rigid
PVC/wood-ﬂour composites, Journal of Vinyl & Additive Technology, 10(4): 179-
186 (2004).

NE. Marcovich, M.M. Reboredo, and M.I. Aranguren, Moisture diffusion in
polyester—woodﬂour composites, Polymer, 40(26): 7313—7320 (1999).

AK. Bledzki and O. F aruk, Creep and impact properties of wood ﬁbre—
polypropylene composites: Inﬂuence of temperature and moisture content,
Composites Science and Technology, 64(5): 693-700 (2004).

Z.A. Mohd Ishak, B.N. Yow, B.L. Ng, H.P.S.A. Khalil, and H.D. Rozman,
Hygrothermal aging and tensile behavior of injection-molded rice husk-ﬁlled

polypropylene composites, Journal of Applied Polymer Science, 81(3): 742-753
(2001)

J. Markarian, Wood-plastic composites need protection, Plastics Engineering, 63(7):
54-56 (2007).

N.M. Stark and L.M. Matuana, Coating WPCs using coextrusion to improve
durability. Proceedings: Coating Wood and Wood Composites: Designing for

Durability, July 23—25, 2007, Seattle, WA.
C. Rauwendaal, Understanding Extrusion, Hanser/Gardner, Cincinnati, (1998).

H.J. Giles Jr., J.R. Wagner Jr., and EM. Mount III, Extrusion: The Deﬁnitive
Processing Guide and Handbook, William Andrew Publishing, New York, (2004).

Z. Giirdal, R.T. Haftka, and P. Hajela, Design and Optimization of Laminated
Composite Materials, John Wiley & Sons, New York, (1999).

103

14.

15.

J.W. Kim, K. Carlborn, L.M. Matuana, and PA. Heiden, Thermoplastic
modiﬁcation of urea-formaldehyde wood adhesives for increased moisture
Resistance, Journal of Applied Polymer Science, 101(6): 4222-4229 (2006).

S. Jin and L.M. Matuana, Improving water resistance of wood-plastic composites
through coextrusion, SPE ANT EC Papers, 66: 19-23 (2008).

104

CHAPTER 4

Wood Plastic Composites Coextruded with Multi-Walled Carbon

Nanotube Filled Rigid PVC Cap Layer

This chapter is slightly modiﬁed from a research paper that was published in the Polymer

International, 59 (5): 648-657 (2010). It is co-authored by Shan Jin and L.M. Matuana.

105

4.1 Abstract

Wood-plastic composites (WPCs) can absorb moisture in a humid environment
due to the hydrophilic nature of the wood in the composites, making products susceptible
to microbial growth and loss of mechanical properties. Coextruding a poly(vinyl chloride)
(PVC)-rich cap layer on a WPC signiﬁcantly reduces the moisture uptake rate, increases
the ﬂexural strength but, most importantly, decreases the ﬂexural modulus compared to
uncapped WPCs. A two-level factorial design was used to develop regression models
evaluating the statistical effects of material compositions and a processing condition on
the ﬂexural properties of coextruded rigid PVC-wood ﬂour composites with the ultimate
goal of producing coextruded composites with better ﬂexural properties than uncapped
WPCs. Material composition variables included wood ﬂour content in the core layer and
carbon nanotube (CNT) content in the cap layer of the coextruded composites, with the
processing temperature proﬁle for the core layer as the only processing condition variable.
Fusion tests were carried out to understand the effects of the material compositions and
processing condition on the ﬂexural properties. Regression models indicated all main
effects and two powerful interaction effects (processing temperature/wood ﬂour content
and wood ﬂour content/CNT content interactions) as statistically signiﬁcant. Factors
leading to a fast fusion of the PVC-wood ﬂour composites in the core layer, i.e. low
wood ﬂour content and high processing temperature, were effective material composition
and processing condition parameters for improving the ﬂexural properties of coextruded
composites. Reinforcing the cap layer with CNTs also produced a signiﬁcant
improvement in the ﬂexural properties of the coextruded composites, insensitive to the

core layer composition and the processing temperature condition.

106

4.2 Introduction

Wood-plastic composites (WPCs) have experienced signiﬁcant market expansion
in recent years as a replacement for solid wood, mainly in outdoor applications such as
railings, decking, landscaping timbers, fencing, playground equipment, window and door
frames, etc. The acceptance of WPCs into the construction industry to replace solid wood

in several outdoor applications contributes to the popularity of WPCs [1].

Originally marketed as weather-resistant products requiring little or no
maintenance compared to solid wood, expectations fail in accounting for exposure to
outdoor environment of ultraviolet rays and moisture which affects the product retaining
its new appearance without maintenance. With many years in outdoor service, customers
have complained of surface staining, discoloration, and other maintenance-oriented issues
in WPC products [1—3]. Studies have shown that WPCs can absorb moisture in a humid
environment due to the highly hydrophilic nature of the wood ﬂour in the composites,
especially at high loading level, making products more prone to microbial growth and
biological decay, thus affecting the appearance of the products [2—6]. Moisture
absorption also reduces the adhesion between the ﬁller and polymer matrix, which results
in modest loss of mechanical properties [7]. Aware of these durability issues, the industry
continues to take steps to ensure the long-term performance of WPC products. In
particular, WPC producers seeking improved products focus on water absorption
resistance, color and appearance retention, physicomechanical property retention, as well

as mould and mildew resistance over time [1—16].

Several proposed approaches substantially reduce or delay the moisture and water

uptake of WPCs. Reports mention the use of performance-enhancing additives such as

107

lubricants, coupling agents and other compatibilizers, encapsulation of exposed wood
during compounding or the use of protective coatings as methods to reduce the moisture
uptake of WPCs [14]. Although proven to reduce moisture absorption of WPCs, these
suggested approaches have not eliminated this factor completely. Unsurprisingly,
investigations have recently demonstrated that covering WPCs with hydrophobic layers
decreases moisture absorption. A cap layer of polyoleﬁn or poly(vinyl chloride) (PVC)
coextruded over a WPC provides improved moisture resistance compared with an

uncapped control WPC [7—19].

Although coextruded WPCs have advantages over uncapped (control) WPCs in
terms of less moisture absorption and a slower rate of moisture uptake, their performance
in terms of ﬂexural properties is very sensitive to the properties of the polymer of the cap
layer [18]. It has been shown that coextruding rigid PVC cap layers on WPCs
signiﬁcantly increases their ﬂexural strength but decreases their ﬂexural modulus
compared to control samples. The increased bending strength and decreased bending
modulus induced by coextruding a PVC layer over a WPC are attributed to the greater

strength but lower stiffness of PVC compared to WPCs [18].

In addition to durability, the mechanical properties of WPCs are another
important concern. As some surveys indicate, key factors to open new markets for WPCs
are good durability and mechanical properties [20]. With ﬂexural properties as critical,
determining the maximum load capacity and span ratings for the acceptance of a material
in various lumber applications, enhancing the strength and stiffness of coextruded WPCs
could not only improve their acceptance in load-bearing structural applications, but also

Open new applications for these products, thus expanding their market share [1].

108

The mechanical properties of coextruded samples depend on the properties of the
materials in both outer and inner layers. Generally, stronger and stiffer materials in the
outer layer enhance the overall performance of the coextruded samples compared to
products made with the inner material alone [21, 22]. Therefore, coextruded WPCs with
stronger and stiffer PVC outer layers could offset the reduction in ﬂexural modulus

observed in previous investigations [18].

Thus far, much effort has been expended on manufacturing PVC products with
satisfactory performance. Notably, signiﬁcant factors include the choice of proper resin,
processing additives and processing conditions in order to achieve optimal mechanical
properties [23—26]. In general, both the processing and performance of PVC products are
intrinsically linked to the resin molecular weight, which is expressed as inherent viscosity
(IV) or k-value, the morphology of the particles (e.g. apparent bulk density, porosity,
particle size distribution), the additives used and other parameters [24]. All factors being
equal, the intrinsic values of many mechanical properties of PVC products increase as the
resin molecular weight increases, similar to other thermoplastics [23, 24]. However,
reports indicate some confusion in the literature while attempting to correlate mechanical
properties to the molecular weight of PVC. It has been shown that increased molecular
weight has little effect on increasing the yield stress but a signiﬁcant effect in decreasing
the tensile modulus of PVC [27]. This results partly because of the dependency of the
resin’s properties on the processing conditions used in the manufacturing of PVC
products, which affect the fusion of PVC [23—26]. Since the resin requires proper fusion
in order to achieve optimal mechanical properties, the higher the molecular weight, the

slower the fusion and the more difﬁcult the processing [23—26, 28]. These ﬁndings

109

suggest that processing conditions such as processing temperature play an important role

in the mechanical performance of PVC products.

Reinforcement of PVC matrices with inorganic and organic ﬁllers constitutes
another approach to improve mechanical properties [29, 30]. Generally, increasing the
amount of ﬁllers in PVC signiﬁcantly enhances its stiffness due to the greater stiffness of
the ﬁllers [30]. This approach also leads to a signiﬁcant reduction in strength due to the
weak interfacial adhesion between the ﬁller and the matrix [30, 31]. However, recent
investigations have shown that PVC ﬁlled with nanoparticles such as nanoclay, carbon
nanotubes (CNTs) and nano-CaCO; possesses both greater ﬂexural strength as well as
modulus than the base resin [32—39], suggesting the potential for enhancement of the

ﬂexural properties of coextruded WPCs by reinforcing the cap layer with nanoparticles.

Consequently, using a two-level factorial design, the study reported here
examined the effects of material compositions in both outer and inner layers as well as a
processing condition on the ﬂexural properties of coextruded rigid PVC/wood ﬂour
composites. Particular emphasis was placed on understanding the effects of processing
temperature for the core layer, reinforcing the outer layer with CNTs and the amount of

wood ﬂour in the core layer on the performance of coextruded composite samples.

110

4.3 Experimental

4.3.1 Materials

Table 4.1 lists the PVC resin and other additives used in the rigid PVC
formulation. A 0.425 mm (40 mesh) maple ﬂour from American Wood Fibers was used
as wood ﬂour. The wood ﬂour was oven-dried at 105°C for 48 h to remove moisture
before processing. Industrial grade 90 wt% multi-walled carbon nanotube (diameter: 10—
30 nm, length: 10—30 pm), purchased from Cheap Tubes (Vermont), served as

reinforcing material for the outer layer.

111

Table 4.1. Formulation of rigid PVC.

 

Ingredients

Content (phr)

 

 

PVC resin (K-value = 66.4) (Shintech)
Tin stabilizer (PlastiStab 2808-Halstab)

Calcium stearate (Synpro)

Parafﬁn wax (Gulf Wax)
Processing aid (Paraloid K-120)l
Processing aid (Paraloid K-175)l

Impact modiﬁer (K-334)l

100
2

1.5

10

 

1Supplied by Rohm & Haas Co.

112

4.3.2 Rigid PVC formulation

The PVC resin and other additives listed in Table 4.1 were blended in a 20 L
high-intensity mixer (Papenmeier, Type TGAHK20) at room temperature for 10 min.
This compounded material was used as matrix in inner and outer layers of coextruded

PVC-wood ﬂour composites.

4.3.3 Compounds for core layer

The formulations used in core layers consisted of blends of PVC and wood ﬂour.
The composites used in the inner layers contained 40 or 60 wt% wood ﬂour based on the
total weight of the composites. These blends were also mixed in a high-intensity mixer at

room temperature for 10 min.

4.3.4 Compounds for outer layer

Outer layers of coextruded WPCs used both unﬁlled and CNT-ﬁlled rigid PVC.
The addition of CNTs into PVC matrix followed the approach described in our previous
study [33]. The pre-blended PVC was placed into a 60 ml electrically heated three-piece
internal mixer/measuring head (3:2 gear ratios) with roller-style mixing blades (C.W.
Brabender Instruments, Inc.) and the mixing process started to follow the fusion curve.
CNTs (5% based on the total weight of the blend) were added into the mixer at the onset
of the fusion, i.e., the point where the material reached a void-free state and started to
melt. At this point, the PVC primary particles were reduced from micrometric in size to
submicrometric particles [32], allowing for a much better dispersion of CNTs into the

matrix. After CNT addition, the materials remained in the mixer until reaching

113

equilibrium torque. The mixing process ran at a constant mixer temperature of 150 °C for
5 min. The rotor speed was 50 rpm and weight charge set at 65 g. The experiments had a

5 kg dead weight on the top of the ram throughout [32, 33, 40].

4.3.5 Experimental design

With multiple variables to account for, as in this study, difﬁculty arises in
studying the system with the common approach of varying only one factor at a time,
while holding the others constant. This approach is not only time-consuming, but it also
presents problems in identifying all of the interactions between factors [41, 42].
Developing a mathematical model allows for greater efﬁciency in investigating these
systems by describing the relationship between the response and independent variables,
and ultimately determining the signiﬁcance of individual factors and multifactor
interactions [43—47]. Response surface methodologies such as a central composite design
(CCD) and a Box-Behnken design (BBD) present versatile methods to statistically model
and optimize response variables affected by multiple independent factors [41, 42].
However, the limitation in processing temperature to avoid thermal degradation of
PVC/wood ﬂour composites makes the use of CCD and BBD methods difﬁcult in this
study to ﬁt the axial points required in these designs. Instead, preference is for a two-

level factorial design since it includes only factorial points in the experiments. The value
of a two-level factorial design lies in its ability to examine the effects of more than one
variable in a system and its ability to reveal the effects of interactions between variables

unobtainable from the traditional method of changing one variable at a time [46, 47].

114

A two-level factorial design was performed using Design Expert® software v.7
(STAT-EASE, Corp., Minneapolis, MN). The three variables studied were the wood ﬂour
content in the inner layer, the CNT content in the outer layer and the processing
temperature for the inner layer, whereas the responses obtained were the ﬂexural strength
and modulus. Table 4.2 lists the experimental design matrix, showing two different
methods of displaying the levels of factors, i.e., factor levels in actual scale (or the actual
values in the experiments) and in coded scale (as +1 and -1 for high and low levels,

respectively). The coded factor levels are deﬁned as follows:

Coded factor levels = Actual value _ Factor mean (Equation 4.1)

(Range of factorial values/2)

 

Each design point included at least six replicates to determine the statistical

signiﬁcance of the results.

After obtaining all responses, the software generated ﬁrst-order surface models
describing the main effects and interactions on the response in terms of coded variables

based on Equation 4.2 [41, 42].

3 3 3

Y: BO+ZBHX+Z 20...XiX

1 (Equation 4.2)
=11 -11_1 J

J

Where Y is the predicted response; X, the main effect, Xin the interaction, [30 the
intercept, and B, and [3,,- the regression coefﬁcients, which are one-half of the
corresponding factor effect estimates. The model functioned as a ﬁrst-order response

surface model because the experiment design did not include the center point.

115

Table 4.2. Factorial design matrix in terms of both actual and coded factor levels.

 

 

 

 

 

 

 

Factors Responses
A: B: C:
Temperature Wood ﬂour CNT Strength Modulus
proﬁle for content in content in (MPa) (MPa)
core core (%) cap (%)
low (-1) 40 (-1) 0 (—1) 56.5 2887.6
high (+1) 40 (-1) 0 (-1) 62.8 3933.6
low (—1) 60 (+1) 0 (--I) 37.2 2243.3
high (+1) 60 (+1) 0 (-1) 50.9 3707.4
low (-1) 40 (-1) 5 (+1) 59.9 3417.3
high (+1) 40 (—1) 5 (+1) 64.7 4276.8
low (—1) 60 (+1) 5 (+1) 48.6 2968.0

high (+1) 60 (+1) 5 (+1) 57.3 4600.7

 

116

4.3.6 Coextrusion of PVC/wood-flour composites

Table 4.2 shows each design point, coextruded in a system consisting of a
coextrusion die connecting a twin screw extruder (TSE) for the inner layer and a single
screw extruder (SSE) for the outer layer (Figure 4.1). Formation of the core layer used a
32 mm conical counter rotating TSE (C.W. Brabender Instruments, Inc.) with a length-to-
diarneter ratio of 13:1. Production of cap layers used a 19 mm SSE (C.W. Brabender
Instruments, Inc.) with a length-to-diameter ratio of 30:1. The speciﬁcally designed die
used in coextrusion insured that the cap layers covered the entire core layer of WPCs
(complete encapsulation of core layer), which created rectangular samples with a nominal

width and depth of 2.54 cm (1 inch) and 0.95 cm (3/8 inch), respectively.

While the temperature proﬁles of the twin screw extruder varied from the heating
zone 1 to zone 3, the temperature proﬁle of the SSE and the die remained constant. The
TSE maintained a rotational speed of 30 rpm throughout the experiments. Conversely,
various rotational screw speeds were used for the SSE to produce coextruded samples
with a cap layer having constant thickness (ca 1 mm). Table 4.3 presents the conditions

used during coextrusion process.

117

 

 

 

 

 

 

 

Twin screw extruder

/ (core layer)

Single screw extruder

 

 

  
 

 

 

 

 

 

 

 

(cap layer)
Die
‘_ Coextruded sample
-— ‘K (Fully capped)
\\
Core layer
Cap layer
Figure 4.1. Schematic of the coextrusion system used in this study.

118

Table 4.3. Processing conditions for coextrusion process.

 

 

 

 

 

 

 

 

 

Die
Variables Processing conditions for extruders temperature
(°C)
. Twin screw extruder
Core layer processmg
temperature Temperature proﬁlel Screw speed
(°C) (rpm)
High 190-185-177 30 173
Low 180-175-167 30 173
_ Single screw extruder
Cap layer material
. . l
compos1tlons Temperature proﬁle Screw speed
(°C) (rpm)
Unﬁlled PVC Cap 155-165-167 9-12 173
5% CNT-ﬁlled
155-165-167 5-7 173

PVC Cap

 

1The temperature proﬁles are shown from the hopper to the die.

119

4.3.7 Flexural property evaluation

Three-point ﬂexural tests were carried out using an Instron 5585H testing
machine equipped with Bluehill 2 software. Tests were performed in a walk-in
conditioning room at 23:1:2 °C and 50i4 % relative humidity. Samples were stored in this
conditioning room for at least 48 h prior to property testing. The dimensions of samples
were about 200X25.4X9.5 mm3 with the tests performed at a crosshead rate of 4.3
mm/min in accordance with ASTM standard D7031-2004. At least six replicates were

tested to obtain an average value for each design point (Table 4.2).

4.3.8 Fusion tests

Performance of fusion tests assisted in understanding the inﬂuences of material
compositions and temperature processing condition on the ﬂexural properties of
coextruded composites. A 7.5 hp Intelli-Torque Plasti-Corder Torque Rheometer® drive
(C.W. Brabender Instruments, Inc.) was employed for fusion studies using a 60 ml
electrically heated three-piece internal mixer/measuring head (3:2 gear ratio) with roller
style mixing blades (C.W. Brabender Instruments, Inc.) [32, 33, 40, 47]. The tests ran at
constant mixing temperatures of 180 and 190 °C for 10 min. All compounds used the
weight charge set at 60 g with the rotor speed maintained at 50 rpm. A 5 kg dead weight
was put on top of the ram throughout the experiments. The Brabender® Mixer Program
(WINMIX, version 3.2.11) recorded the fusion time. Each formulation included at least

three replicates to obtain average values for the fusion characteristics.

120

4.3.9 Scanning electron microscopy

SEM images were obtained to qualitatively view the effect of processing
temperatures (180 and 190 °C) on the extent of fusion inside the PVC ﬁlled with 40 and
60% wood flour (core layer). A JEOL JSM-6400 SEM at 15kv accelerating voltage was
used for SEM images. The samples were oven-dried at 105°C to remove the moisture and

gold coated before observation.

121

4.4 Results and Discussion

4.4.1 Statistical signiﬁcance of effects

A two-level factorial design was employed to evaluate the statistical effects of
processing temperature for the core layer (factor A), wood ﬂour content in the core layer
(factor B) and CNT content in the cap layer (factor C) on the ﬂexural properties of
coextruded rigid PVC/wood ﬂour composites. The analysis of the data assists in
identifying the major factors affecting the ﬂexural properties as well as understanding

interactions between different variables.

For a two-level factorial design, the normal probability plot, a graphical tool,
assesses the importance of the factor effects. Figure 4.2 shows the normal probability
plots of the effect estimates for ﬂexural strength and stiffness of coextruded composites.
While all negligible and non-signiﬁcant effects lie along the straight line, all the main
effect factors (A, B and C) and the most signiﬁcant two-factor interaction terms (BC for
ﬂexural strength and AB and BC for ﬂexural modulus) affecting the ﬂexural properties

fall off to the left (negative effect) or to the right (positive effect) of the line.

122

 

 

 

 

 

503m
00...“ F N. mmde mmdmv hNde mhdmw-
_ _ L _ r
m- 1
0m —
m<u - m1
0- .1.
(I I
«GE .0852

365 E256 kzo H0
36V Eoucoo .50: 0003 um
2:05 BEEeQEOP u<

Amt—2V 03.2002
.05 memxw120ﬁwwo

P

OLD
‘-

mo
00’

O)
0)

4111199de % neutron

3V

56¢on vwdommp 3.3mm-

P

_

£38888
505100050 >m Ewt @3558 m0 2:309: :2:on 3v Ea ﬁwcobm .830: A3 00% 300% “.0 $03 bzﬁunoa 3882

80cm

wmdmvwm. mmdawmm.
_ _ _

 

 

I
I

 

mo
030)

0
<
IIIIIIIIIIWWWWWWIHHIIII
O O O 8 O O to
Q N In N 1-

I
O)
O)

 

 

.03 .0552

33 E628 #20 Ho
36. E28 50: 803 Um
050...... 0.280050... 3

3.82003 50:25.
“0.0 ...mmaxmigmmo

MIIQPQOJd 881mm

:3

aseam

123

The perturbation plot, another tool, assists in comparing the relative signiﬁcance
of the main effects on the response. It illustrates the changes in a response as each factor
moves from the chosen reference, with all other factors held constant at the middle level
of the design space [41, 42]. Because this study used the processing temperature (factor A)
as a categorical factor, the perturbation plots for ﬂexural strength and modulus were
created at both low and high levels of processing temperature for the core layer (Figures
4.3 and 4.4) while keeping the wood flour content and CNT content at their middle level
of the design space. The midpoint acts as the reference point (the coded zero level of each
factor). Steep slopes observed for the main factors B and C in Figure 4.3 indicate the
sensitivity of the ﬂexural strength to both wood ﬂour content in the core layer (factor B)
and CNT content in the cap layer (factor C), regardless of the processing temperature
used for the core layer. While the CNT content in the outer layer has a signiﬁcant impact
on the stiffness of coextruded composites independent of the processing temperatures
used for the core layer, the amount of wood ﬂour in the core layer largely affects the
stiffness of the coextruded composites at low processing temperature (Figure 4.4(a)) and

shows only a small effect at high processing temperature (Figure 4.4(b)).

124

Perturbation

 

 

 

 

152761-1
(a)
to 1182900
“2 B
N
5‘
’c?
a. C
E 83819—
‘5»
C
0 C
a
a)
V 49340-a
143774
I 1 I T I
4000 0500 0.000 0.500 1000

Deviation from Reference Point

Figure 4.3. Perturbation plots of the ﬂexural strength of coextruded composites
against wood ﬂour content in core layer and CNT content in cap layer of
coextruded composites processed at (a) low temperature and (b) high

temperature proﬁle for the core layer.

125

 

 

 

 

Perturbation
1527614
B (b)
C
8 11829)—
N’ C
5‘
it?
O.
E 83819-
:E
O)
c
(D
.b
(I)
V 49348-
14877ﬁ
l | F l l
1 (XI) -0 500 0.000 0 5C!) 1 000
Deviation from Reference Point
Figure 4.3. Perturbation plots of the ﬂexural strength of coextruded composites

against wood ﬂour content in core layer and CNT content in cap layer of
coextruded composites processed at (a) low temperature and (b) high

temperature proﬁle for the core layer.

126

Modulus (MPa)

Figure 4.4.

Perturbation

 

 

 

 

4912 .2 (a)
4190-4
3468—

B C
2746— C B
2023—

I l T I l

-10 -05 0.0 0.5 1.0

Deviation from Reference Point

Perturbation plots of the ﬂexural modulus of coextruded composites
against wood ﬂour content in core layer and CNT content in cap layer of
coextruded composites processed at (a) low temperature and (b) high

temperature proﬁle for the core layer.

127

Modulus (MPa)

Figure 4.4.

Perturbation

 

 

 

 

 

4912—-
(b)
C
4190- B M
/
3468—
2746—
2023—
I T I T j
-1 o -o 5 0.0 0.5 1 0

Deviation from Reference Point

Perturbation plots of the ﬂexural modulus of coextruded composites
against wood ﬂour content in core layer and CNT content in cap layer of
coextruded composites processed at (a) low temperature and (b) high

temperature proﬁle for the core layer.

128

The analysis of variance (ANOVA) conﬁrms the interpretation of the data based
on the normal probability plots (Figure 4.2) and perturbation plots (Figures 4.3 and 4.4)
according to the p-values (Prob > F) showing the statistically signiﬁcant factor effects

(Tables 4.4 and 4.5).

A regression analysis was carried out to obtain the best-ﬁt model to the
experimental data. The analysis suggests that a power model with positive lambda value
best ﬁts the relationship between the ﬂexural strength and the three independent variables,
whereas a linear model best ﬁts the ﬂexural modulus. The following equations describe
the derived regression models predicting the ﬂexural strength and stifﬁiess of coextruded

composites in terms of coded factors:

(Strength)2'85 = +95037.5 +18542.3A — 27846.68 + 12025.2(: + 4179.8BC (Equation 4.3)

Modulus = +3523.5 + 606.2A —119.9B + 292.2C +144.3AB + 88.5BC (Equation 4.4)

where A is the processing temperature for the core layer, B the wood ﬂour content in the

inner layer and C the CNT content in the outer layer of coextruded composites.

129

Table 4.4.

ANOVA for selected factorial model of ﬂexural strength of coextruded

 

 

composites.
Source Sum of Degree Mean F- Prob > Fa
squares of square Value
freedom
Model 6.150><10'0 4 1.538><1010 156.51 < 0.0001
A 1.650x10IO 1 1.650><10lo 167.99 < 0.0001
B 3.722x10'0 1 3.722><10'0 378.87 < 0.0001
C 6.941 ><109 1 6.941><109 70.65 < 0.0001
BC 8.386x108 1 8.386x108 8.54 0.0055
Lack of ﬁt" 4.019><108 1 1.34><108 1.40 0.2563

_

 

a
Values of Prob > F less than 0.0500 indicate model terms are signiﬁcant. In this case A,

B, C, and BC are the signiﬁcant model terms.

The “Lack of Fit” F value of 1.4 implies the lack of ﬁt is not signiﬁcant relative to the

pure error. There is a 25.63% chance that a lack of ﬁt F value this large could occur due

to noise. Non-signiﬁcant lack of ﬁt is good.

130

Table 4.5. ANOVA for selected factorial model of ﬂexural modulus of coextruded

 

 

composites.
Source Sum of Degree of Mean F- Prob > Fa
squares freedom square Value
Model 2.380x107 5 4.760><106 103.58 < 0.0001
A 1.764><107 1 1.764><107 383.78 < 0.0001
B 6.897><105 1 6.897x105 15.01 0.0004
C 4.099x106 1 4.099><106 89.20 < 0.0001
AB 9.993 x 105 1 9.993><10S 21.75 < 0.0001
BC 3.760x105 1 3.76O><105 8.18 0.0066
Lack of ﬁtb 1.289x105 2 64434.78 1.43 0.2511

 

aValues of Prob > F less than 0.0500 indicate model terms are signiﬁcant. In this case

A, B, C, and BC are the signiﬁcant model terms.

bThe “Lack of Fit” F value of 1.43 implies the lack of ﬁt is not signiﬁcant relative to

the pure error. There is a 25.11% chance that a lack of ﬁt F-value this large could

occur due to noise. Non-signiﬁcant lack of ﬁt is good.

131

With Equations (4.3) and (4.4) displayed in terms of coded factors, the absolute
value of regression coefﬁcients indicates the relative signiﬁcance of the effect of factors
on the response, and the algebraic sign of each factor reﬂects the positive or negative
relation between factors and response. As shown in these equations, all signiﬁcant main
effect factors (A, B and C) and two-factor interaction terms (AB and BC) have the same
algebraic signs. These expected results occur because with the ﬂexural strength as the
highest stress experienced within the material at its moment of rupture, and the stiffness
as the ratio of stress to strain in ﬂexural deformation, the effects would inﬂuence both
strength and stiffness in a similar way [48]. In other words, the effects with a positive

impact on strength will also have a positive impact on stiffness and vice versa.

Because factors A and C have positive algebraic signs, an increase in processing
temperature for the core layer (factor A) and CNT content in the cap layer (factor C)
improves the ﬂexural properties of coextruded composites. By contrast, the negative
effect of wood ﬂour content (factor B) suggests that increasing the wood ﬂour content
from the low level (40%) to the high level (60%) worsens the ﬂexural properties of
coextruded composites. The amount of wood flour in the core layer of coextruded
composites (factor B) has the largest effect on ﬂexural strength (large regression
coefﬁcient), followed by the processing temperature (factor A) and CNT content (factor
C). A different trend is obtained for ﬂexural modulus: largely affected by the processing
temperature for the core layer followed by the amounts of CNT in the outer and wood
ﬂour in the inner layers. Equations (3) and (4) also reveal powerful interactions between

processing temperature and wood ﬂour content (interaction AB), as well as wood ﬂour

132

content and CNT content (interaction BC). These interactions have positive algebraic

signs, indicating a positive effect on ﬂexural properties of the coextruded composites.

With three of the main effect factors (A, B and C) that affect the ﬂexural
properties of coextruded composites being part of two signiﬁcant interactions (AB and
BC), the co-dependent behavior of one factor on the level of the other would make
investigating these main factors separately inappropriate [43-47]. Consequently, the

signiﬁcant interactions (AB and BC) require simultaneous consideration.

4.4.2 Effect of the interaction factor on the ﬂexural strength

Examining the two-dimensional response surface plots generated according to
Equations (3) and (4) would assist in understanding the relationship between the ﬂexural
properties of coextruded rigid PVC/wood ﬂour composites and the signiﬁcant main effect
factors as well as their interactions. Figure 4.5 shows the interaction graphs displaying
ﬂexural strength dependence as a function of the wood ﬂour content in the inner layer
and CNT content in the outer layer interaction (interaction BC). This shows the
interaction’s effect for the processing temperatures for the core layer set at low (Figure

4.5(a)) and high levels (Figure 4.5(b)).

133

 

152761—1
(a)

 

 

 

:3 "329°" 5% CNT
(\i
<
a ./
O.
a same—I
5
0:
C
0
1.. /
(‘0’ 49348—
0°/o CNT
14877—1
—I I I I I
40 45 50 55 60

8: Wood flour content (%)

Figure 4.5. Interaction graphs of the variation in ﬂexural strength as a function of the
interaction between wood ﬂour content in the core and CNT content in the
cap layers of coextruded composites processed at (a) low temperature and

(b) high temperature proﬁle for the core layer.

134

 

 

 

 

152781—I
5% CNT (b)
ID 118290—
00
N
it
67"; /
2 83819__ O% CNT
5
U,
C
G)
.b
(D
v 49348—
14877—
1 1 I 1 l
40 45 50 55 60

B: Wood flour content (%)

Figure 4.5. Interaction graphs of the variation in ﬂexural strength as a function of the
interaction between wood ﬂour content in the core and CNT content in the
cap layers of coextruded composites processed at (a) low temperature and

(b) high temperature proﬁle for the core layer.

135

Observations show a dependency of the ﬂexural strength on the wood ﬂour
content and CNT content. As previously mentioned, the ﬂexural strength of coextruded
composites decreases with increasing wood ﬂour content, but increases with increasing
CNT content. The addition of CNT in the outer layer produces more notable changes in
the ﬂexural strength; uninﬂuenced by the core layer composition and the processing
temperature condition (Figure 4.5). This expected trend is attributed to the reinforcing
effect of CNTs [33]. However, the amount of CNTs in the outer layer has a minimal
effect on the ﬂexural strength of coextruded composites at low wood ﬂour content (40%)
but its effect intensiﬁes with increased wood ﬂour content (60%) in the core of
coextruded composites. Comparing the difference in magnitude between the lines for
unﬁlled (0% CNT) and ﬁlled (5% CNT) cap layers in Figure 4.5 veriﬁes this effect.
Additionally, non-parallel lines in Figure 4.5 reﬂect the interaction effect on the ﬂexural
strength. Since the strength of WPCs decreases as the amount of wood ﬂour increases in
the composites [30], reinforcing the cap layer with CNTs has a greater effect on the
strength of coextruded composites with a greater amount of wood ﬂour in the core layer.
The shallow slope of C+ (5% CNT) and the steep slope of C— (0% CNT) imply that
CNT-reinforced cap layer offsets part of the negative effect that high wood ﬂour content

in the inner layer has on the strength of coextruded composites.

136

 

Evidence indicates that increasing the processing temperature for the core layer
signiﬁcantly increases the ﬂexural strength of the coextruded composites, insensitive to
the material composition (Figure 4.5). Generally, strength and stiffness of individual
components (rule of mixtures), as well as the interface’s ability to transfer stresses
between the filler and matrix, determine the strength and stiffness of polymer composites
[30, 31, 49-52]. The advantage of processing at a high temperature lies in its effect on the
fusion behavior of PVC matrix [23-26]. The degree of fusion in PVC affects not only the
interfacial adhesion, but also the stress-transferring ability between components in the

composites [2, 40].

137

 

Table 4.6. Effects of wood ﬂour content and test temperature on the fusion time of

rigid PVC/wood-ﬂour composites.

 

Wood ﬂour content in Fusron time (S)

 

 

 

 

composites (%) 180oC 190°C
40 48 34
60 3 1 1 21 7

 

138

As shown in Table 4.6, the fusion time of PVC/wood ﬂour composites remains
highly sensitive not only to the processing temperature for the core layer, but also to the
amount of wood ﬂour in the inner layer. A slight increase in processing temperature
(10°C) reduces the fusion time signiﬁcantly (about 30%), which promotes a high fusion
level in the WPC used in the core layer. Increasing the processing temperature reduces
the fusion time, mainly because the interdiffusion between PVC microparticles forming a
three-dimensional network among polymer chains proceeds at a much faster rate [53].
The formation of this three dimensional network would proceed at a slower rate with a
lower processing temperature. PVC particles well fUSCd at elevated processing
temperature indicate an enhancement in the wood ﬂour’s encapsulation by the matrix, as
well as the adhesion between the WPC in the core layer and the PVC in the cap layer,
leading to improved ﬂexural strength of coextruded composite samples. The SEM images
shown in Figure 4.6 support this contention. While the surfaces of PVC/wood ﬂour
composites processed at a high temperature are relatively smooth, those processed at a

reduced temperature have relatively coarse surfaces.

139

 

Figure 4.6.

.7. 500 pm

SEM micrographs of the rigid PVC/wood-ﬂour composites core layer

containing various amounts of wood ﬂour and processed with different
temperature proﬁles: (a) 40% wood ﬂour/high temperature; (b) 40% wood
ﬂour/low temperature; (c) 60% wood ﬂour/high temperature; (d) 60%

wood ﬂour/low temperature.

140

Similarly, increasing wood ﬂour content in the composites generates a prolonged
fusion time (Table 4.6). The poor thermal conductivity of wood may attribute to the
delayed fusion time in the composites, which decreases the transfer of heat and shear
throughout the PVC grains [40]. Considering the corresponding increase in the fusion
time with increasing wood ﬂour content, the composites with high wood content (60%)
would require additional thermal energy to be absorbed by the composite in order for the
PVC particles to fuse together. The lower bulk density of the composites could also
explain the delayed fusion process in the composites with high wood content. Fusion
begins once a sufﬁcient amount of heat has been put into the resin grains [23-26]. A high
bulk density creates ample compaction, additional shear and accelerated heat build-up
[54]. Because the mass fraction of the resin in the composites decreases as the wood ﬂour
content increases, the bulk density of neat resin will decrease as the amount of wood
increases in the composites {40]. Consequently, for the equivalent mixer charge weight
(mass of the material), the build-up of heat will be slow in the composites with high
wood ﬂour content, which delays the fusion process. As mentioned, slow fusion results in
poor adhesion between wood ﬂour and PVC. Moreover, SEM micrographs of samples
containing 40 and 60% wood ﬂour support these ﬁndings. As seen in Figures 4.6 (a) and
(b), a relatively smooth cross-section surface occurs in PVC/wood ﬂour core with low
wood ﬂour content compared to relatively rough surfaces occurring with those of high
wood ﬂour content (Figures 4.6 (c) and (d)) at the same processing temperature.
Smoother surfaces suggest a well-fused PVC matrix, better encapsulation of wood ﬂour
by the PVC matrix in the core layer and good interfacial adhesion between the WPC in

the core and the PVC matrix in the cap layers.

141

4.4.3 Effect of the interaction factors on the ﬂexural modulus

Unlike the ﬂexural strength (Equation (4.3)), the two-level factorial design reveals
two powerful interactions (interactions AB and BC) positively affecting the ﬂexural
modulus of coextruded composites (Equation (4.4)). However, examining the relationship
between the ﬂexural modulus and these interactions involves considering only the
interaction between processing temperature and wood ﬂour (interaction AB), due to its
predominant effect on the ﬂexural modulus compared to the wood ﬂour content/CNT

content interaction (interaction BC).

Figure 4.7 shows the interaction graphs of the variation of the ﬂexural modulus as
a function of the processing temperature for the core layer and wood ﬂour content in the
inner layer (interaction AB). The ﬁgure incorporates the effect of this interaction for
coextruded composites with unﬁlled (0% CNT) (Figure 4.7(a)) and 5% CNT-ﬁlled

(Figure 4.5(b)) rigid PVC cap layers. The data demonstrate two distinct trends.

First, the ﬂexural modulus of coextruded composites does not conform to the rule
of mixtures with the core layer processed at low temperature (Figure 4.7). Although the
modulus of PVC/wood ﬂour composites increases with increasing wood ﬂour content
(rule of mixtures) [29-31], contrarily, this study found the opposite trend (Figure 4.7).
Coextruded composites with a large amount of wood ﬂour (60%) have lower ﬂexural
modulus than their counterparts with lower wood ﬂour content (40%), regardless of the
cap layer composition. Prolonged fusion time and inadequate interfacial adhesion
induced by the low processing temperature proﬁle may account for the decreased
stiffness. However, the CNT-reinforced cap layer signiﬁcantly improves ﬂexural stiffness,

regardless of core layer composition.

142

Modulus (MPa)

Figure 4.7.

 

4912— (a)

4190 —

    

40% wood ﬂour

\

 
    

 

 

 

3468 —
2746 — '\
60% wood flour
2023 -
l 1
'0" high

A: Temperature profile

Interaction graphs of the variation in ﬂexural modulus of coextruded
composites as a function of the processing temperature for core layer-
wood ﬂour content in inner layer interaction: (3) no CNT in cap layer and

(b) cap layer reinforced with 5% CNT.

143

Modulus (MPa)

Figure 4.7.

 

 

    
    

4912 —I
( b)
4190 -
40% wood ﬂour
3453 _
2746 -— \
60% wood ﬂour
2023 —

 

 

 

A: Temperature profile

Interaction graphs of the variation in ﬂexural modulus of coextruded
composites as a function of the processing temperature for core layer-
wood ﬂour content in inner layer interaction: (a) no CNT in cap layer and

(b) cap layer reinforced with 5% CNT.

144

Second, processing at an elevated temperature has a noticeable effect on the
ﬂexural modulus of the coextruded composites. As seen in Figure 4.7, its actions in
shortening PVC fusion time and enhancing the interfacial adhesion between wood ﬂour
and PVC matrix in the core layer result in the ﬂexural modulus of coextruded composites
increasing with processing temperature. Nevertheless, the composites with unﬁlled cap
layer do not conform to the rule of mixtures. By contrast, the reinforcing effect of CNTs
results considering their addition in the outer layer produces notable changes in the
ﬂexural strength of coextruded composites, as well as the coextruded composites with
high wood ﬂour content (60%) outperforming the counterparts with 40%wood ﬂour in

the inner layer (Figure 4.7(b)).

Because the ﬂexural properties of coextruded composites depend on the
properties of the components in cap and core layers, the results of this study indicate that
the production of coextruded composites with greater ﬂexural strength and modulus
requires the use of CNT-reinforced PVC in the outer layer combined with a high
processing temperature for the WPC core layer. A low level of wood ﬂour (40%) in the
core layer is beneﬁcial in enhancing the ﬂexural strength of coextruded composites.
Conversely, greater ﬂexural modulus occurs in coextruded composites with high wood

ﬂour content (60%) in the core layer.

As discussed, coextruding the rigid PVC cap layer on WPCs signiﬁcantly
increases their ﬂexural strength but decreases their ﬂexural modulus owing to the greater
strength but lower stiffness of PVC compared to WPCs. The results summarized in Table
4.7 clearly support these ﬁndings. Signiﬁcant improvement in bending strength occurs in

WPCs coextruded with an unﬁlled rigid PVC cap layer. By contrast, the ﬂexural modulus

145

of WPCs is signiﬁcantly reduced when encapsulated with an unﬁlled PVC cap layer. The
greater strength but lower stiffness of unfilled PVC compared to WPCs justiﬁes these
foreseeable results. The reinforcing effect of CNTs is evident since their addition into the

outer layer increases the ﬂexural modulus of coextruded composites.

146

Table 4.7. Comparison of ﬂexural properties between uncapped (control) and

coextruded rigid PVC/wood-ﬂour composites. Composites contained 50%

 

 

 

 

 

wood ﬂour.
F lexural properties
Types of cap layer Strength Modulus
(MPa) (MPa)
None (control) 44.24 :1: 2.31 4518 d: 224
Unﬁlled rigid PVC 54.75 :1: 3.03 3842 :1: 236
5% CNT-ﬁlled rigid PVC 65.04 i 0.89 4915 i 237

 

147

4.4 Conclusions

Coextruding a rigid PVC cap layer on WPCs signiﬁcantly increases their ﬂexural
strength yet decreases their ﬂexural modulus as a result of the greater strength but lower
stiffness of PVC compared to WPCs. An implemented two-level factorial design
developed regression models evaluating the statistical effects of material compositions
and processing conditions on the ﬂexural properties of coextruded rigid PVC/wood ﬂour
composites, with the intention to produce coextruded composites with better ﬂexural
properties than uncapped WPCs. Material composition variables were wood ﬂour content
(40 and 60 %) in the core layer and CNT content (0 and 5 %) in the cap layer of
coextruded composites with the processing temperature proﬁle (low and high) for the
core layer as the only processing condition variable. Fusion tests assisted in
understanding the effects of material compositions and processing condition on the

ﬂexural properties. The conclusions drawn are as stated below.

Regression models indicated all main effects and several powerful interaction
effects (processing temperature/wood ﬂour content and wood ﬂour content/CNT content

interactions) as statistically signiﬁcant.

The ﬁision behavior of PVC/wood ﬂour composites used in the core layer of
coextruded composites, which depends on both processing temperature and wood ﬂour
content, plays a key role in the performance of the coextruded composites. While high
processing temperature accelerated the fusion, increasing the wood ﬂour content in the
composites prolonged fusion time. Rapid fusion resulted in an improved interface
between the PVC matrix and wood ﬂour in the core layer while simultaneously

promoting desirable adhesion between the WPC inner and rigid PVC outer layers of

148

coextruded composites, resulting in better overall ﬂexural properties. These results imply
that factors leading to a fast fusion of PVC/wood ﬂour composites in the core layer are
effective processing and material composition parameters for improving the ﬂexural
properties of coextruded composites. The noticeable reinforcing effect of CNTs included
in the outer layer produced drastic changes in the ﬂexural strength and modulus of
coextruded composites, insensitive to the core layer composition and the processing

temperature condition.

Because the ﬂexural properties of coextruded composites depend on the
properties of the components in cap and core layers, the results of this study indicate that
production of coextruded composites with enhanced ﬂexural strength and modulus
requires the use of CNT-reinforced PVC in the outer layer combined with a WPC core
layer processed at high temperature. A low level of wood ﬂour in the core layer is
beneﬁcial in enhancing the ﬂexural strength, while greater ﬂexural modulus occurs in

coextruded composites with high wood ﬂour content in the core layer.

149

 

4.5 Acknowledgements

The authors gratefully acknowledge the USDA-CSREES Grant-Advanced
Technology Applications to Eastern Hardwood Utilization (Grant No. 2008-34158-19510)
for the ﬁnancial support for this research work. The authors are grateful for the generous
donation of materials provided by the manufacturers listed in the Table 4.1. The authors
gratefully acknowledge Ms. Valerie Edge, an undergraduate student at Michigan State

University, for her thoughtful and constructive comments on the paper.

150

4.6 References

10.

ll.

12.

L.M. Matuana, Recent research developments in wood plastic composites, Journal
of Vinyl & Additive Technology, 15(3): 136-138 (2009).

J.M. Pilarski and L.M. Matuana, Durability of wood ﬂour-plastic composites
exposed to accelerated freeze-thaw cycling. Part 1. Rigid PVC matrix, Journal of
Vinyl & Additive Technology, 11(1): 1-8 (2005).

J.M. Pilarski and L.M. Matuana, Durability of wood ﬂour-plastic composites
exposed to accelerated freeze-thaw cycling. II. High density polyethylene matrix,
Journal of Applied Polymer Science, 100(1): 35-39 (2006).

B. Dawson-Andoh, L.M. Matuana, and J. Harrison, Susceptibility of high density
Polyethylene/wood-ﬂour composite to mold discoloration, Journal of the Institute
of Wood Science, 17(2): 114-119 (2005).

B. Dawson-Andoh, L.M. Matuana, and J. Harrison, Mold susceptibility of rigid
PVC/wood-ﬂour composites, Journal of Vinyl & Additive Technology, 10(4): 179-
186 (2004).

B. Dawson-Andoh and L.M. Matuana, High density polyethylene-wood ﬂour
composite lumber: Efﬁcacy of two proprietary biocides in the control of fungal
colonization and discoloration, Holz als Roh-und Werskstoﬂ: 65(5): 331-335 (2007).

N.M. Stark, Inﬂuence of moisture absorption on mechanical properties of wood
ﬂour-polypropylene composites, Journal of Thermoplastic Composite Materials,
14(5): 421-432 (2001).

J.S. Fabiyi, A.G. McDonald, M.P. Wolcott, and PR. Griffiths, Wood plastic
composites weathering: visual appearance and chemical changes, Polymer
Degradation and Stability, 93(8): 1405-1414 (2008).

N.M. Stark and L.M. Matuna. Characterization of weathered wood-plastic
composite surfaces using FTIR spectroscopy, contact angle, and XPS, Polymer
Degradation and Stability, 92(10): 1883-1890 (2007).

N.M. Stark and L.M. Matuana, Inﬂuence of Photostabilizers on Wood F lour-HDPE
Composites Exposed to Xenon-Arc Radiation with and without Water Spray,
Polymer Degradation and Stability, 91(12): 3048-3056 (2006).

N.M. Stark, L.M. Matuana, and CM. Clemons, Effects of processing method on
surface and weathering characteristics of wood-ﬂour/I-IDPE composites, Journal of
Applied Polymer Science, 93(3): 1021-1030 (2004).

N.M. Stark and L.M. Matuana, Surface chemistry and mechanical property changes
of wood-ﬂour/high density polyethylene composites after accelerated weatherrng,
Journal of Applied Polymer Science, 94(6): 2263-2273 (2004).

151

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

NM. Stark and L.M. Matuana, Surface chemistry changes of weathered
HDPE/wood-ﬂour composites studied by XPS and F TIR' spectroscopy, Polymer
Degradation and Stability, 86(1): 1-9 (2004).

N.M. Stark and L.M. Matuana, Ultraviolet weathering of photostabilized wood-
ﬂour-ﬁlled high density polyethylene composites, Journal of Applied Polymer
Science, 90(10): 2609-2617 (2003).

L.M. Matuana and DP. Kamdem, Accelerated ultraviolet weathering of
PVC/wood-ﬂour composites, Polymer Engineering and Science, 42(8): 1657-1666
(2002)

L.M. Matuana, D.P. Kamdem, and J. Zhang, Photoageing and stabilization of rigid
PVC/wood-ﬁber composites, Journal of Applied Polymer Science, 80(11): 1943-
1950 (2001).

N.M. Stark and L.M. Matuana, Coating WPCs using co-extrusion to improve
durability, Proceedings: Coating Wood and Wood Composites: Designing for
Durability, July 23-25, 2007, Seattle, WA.

S. Jin and L.M. Matuana, Coextruded PVC/wood-ﬂour composites with WPC cap
layers, Journal of Vinyl & Additive Technology, 14(4): 197-203 (2008).

S. Jin and L.M. Matuana, Improving water resistance of wood plastic composites
Through coextrusion, SPE AN TEC Papers, 66: 19-23 (2008).

G. Pritchard, Second-generation wood composites: the US shows Europe the way,
Reinforced Plastics, 49(6): 34-35 (2005).

C. Rauwendaal, Understanding Extrusion, Gardner Publications, Inc., Cincinnati,
(1998)

Z. Giirdal, R.T. Haﬁka and P. Hajela, Design and Optimization of Laminated
Composite Materials, John Wiley & Sons, Inc., New York, (1999).

W.V. Titow, Elementary Principles of PVC Formulation, in PVC T echnology,.
Elsevier Applied Science Publishers, New York, (1984).

P. Kroushl, Handbook of Vinyl Formulating, 2nd Edition, John Wiley & Sons, Inc.,
Hoboken (2008).

J.W. Summers, PVC Handbook, Hanser Gardner Publications, Cincinnati, pp 3-4
(2005)

G. Matthews, Properties of Vinyl Chloride Polymers and Copolymers, in PVC:
Production, Properties and Uses, The Institute of Materials, London, pp 55-56
(1996)

J.C. Garcia and A. Marcilla, A inﬂuence of the type of resin in the gelation and
fusion processes of PVC plastisols. Polymer, 39(2): 431-435 (1998).

152

. [kW—j

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

M. Moghri, H. Garmabi and M. Akbarian, Effect of processing parameters on
fusion and mechanical properties of a twin-screw extruded rigid PVC pipe, J Vinyl
Addit T echnol, 9(2): 81-89 (2003).

L.M. Matuana, C.B. Park, and J .J . Balatinecz, The effect of low levels of plasticizer
on the rheological and mechanical properties of polyvinyl chloride/newsprint-ﬁber
composites, Journal of Vinyl and Additive Technology, 3(4): 265-273 (1997).

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

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

L.M. Matuana, Rigid PVC/(Layered Silicate) nanocomposites produced through a
novel melt-blending approach, Journal of Vinyl & Additive Technology, 15(2): 77-
86 (2009).

O. Faruk and L.M. Matuana, Reinforcement of rigid PVC/wood-ﬂour composites
with multi-walled carbon nanotubes, Journal of Vinyl & Additive Technology, 14(2):
60-64 (2008).

D. Wang and CA. Wilkie, Preparation of PVC-clay nanocomposites by solution
blending, Journal of Vinyl & Additive Technology, 8(2): 238-245 (2002).

C. Wan, X. Qiao, Y. Zhang and Y. Zhang, Effect of different clay treatment on
morphology and mechanical properties of PVC-clay nanocomposites, Polymer
Testing, 22(4): 453-461 (2003).

CH. Chen, C.C. Teng, M.S. Tsai and PS. Yen, Preparation and characterization of
rigid poly(vinyl chloride)/MMT nanocomposites. II. XRD, morphological and
mechanical characteristics, Journal Polymer Science Part B Polymer Physics,
44(15): 2145-2154 (2006).

T. Ren, J. Yang, Y. Huang, J. Ren and Y. Liu, Preparation, characterization, and
properties of poly(vinyl chloride)/organophilic-montrnorillonite nanocomposites
Polymer Composites, 27(1): 55-64 (2006).

D. Wang, D. Parlow, Q. Yao and CA. Wilkie, Melt blending preparation of PVC-
sodium clay nanocomposites, Journal of Vinyl & Additive Technology, 8(2): 139-
150(2002)

C. Xiong, S. Lu, D. Wang, L. Dong, D.D. Jiang and Q. Wang, Microporous
polyvinyl chloride: novel reactor for PVC/CaCO3 nanocomposites, Nanotechnology,
16(9): 1787-1792 (2005).

L.M. Matuana and J .W. Kim, Fusion characteristics of wood-ﬂour ﬁlled rigid PVC
by torque rheometry, Journal of Vinyl & Additive Technology, 13(1): 7-13 (2007).

153

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

RH. Myers and DC. Montgomery, Response Surface Methodology: Process and
Product Optimization Using Designed Experiments, Wiley, New York, pp 91-102
(1995)

DC. Montgomery, Design and Analysis of Experiments, 5“1 Edition, Wiley, New
York, (2001 ).

K. Carlbom and L.M. Matuana, Modeling and optimization of forrnaldehyde-free
wood-based composites using a box-behnken design, Polymer Composites, 27(5):
497-503 (2006).

L.M. Matuana and Q. Li, Statistical modeling and response surface optimization of
extruded HDPE/wood-ﬂour composite Foams, Journal of Thermoplastic Composite
Materials, 17(2): 185-199 (2004).

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

L.M. Matuana and Q. Li, A factorial design applied to the extrusion foaming of
polypropylene/wood-ﬂour composites, Cellular Polymers, 20(2): 115-130 (2001).

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

J.M. Hodgkinson, Mechanical Testing of Advanced Fibre Composites, Woodhead
Publishing, (2000)

D.N. Saheb and JP. Jog, Natural ﬁber polymer composites: A review, Advanced
Polymer Technology, 18(4): 351-363 (1999).

KB. Adhikary, S. Pang and MP. Staiger, Dimensional stability and mechanical
behaviour of wood—plastic composites based on recycled and virgin high-density
polyethylene (HDPE), Composites Part B.“ Engineering, 39(5): 807-815 (2008).

ST. Georgopoulos, P.A. Tarantili, E. Avgerinos, A.G. Andreopoulos and E.G.
Koukios, Thermoplastic polymers reinforced with ﬁbrous agricultural residues,
Polymer Degradation and Stability, 90(2): 303-312 (2005).

J.E. Crespo, L. Sanchez, F. Parres and J. Lopez, Study of the mechanical and
morphological properties of plasticized PVC composites containing rice husk ﬁllers,
Journal of Thermoplastic Composite Materials, 22(3): 259-272 (2009).

E]. Comeaux, CH. Chen, J .R. Collier and RD. Wesson, Fusion study of polyvinyl
chloride (PVC): Relation of Processing time and processing temperature to the
degree of fusion, Polymer Bulletin, 33(6): 701-708 (1994).

JD. Bower, The function of lubricants in processing of rigid poly(vinyl chloride),
Journal of Vinyl & Additive Technology, 8(4):179-182 (1986).

154

 

 

CHAPTER 5

Ultraviolet Weathering of HDPE/Wood-Flour Composites

Coextruded with a Clear HDPE Cap Layer

This chapter was submitted to Polymer Degradation and Stability in June 2010.

155

5.1 Abstract

This study examined the effect of coextruding a clear HDPE cap layer onto
HDPE/wood-ﬂour composites on the discoloration of coextruded composites exposed
to accelerated UV tests. Chroma meter, FTIR-ATR, XPS, SEM, and UV vis
measurements accounted for the analysis of discoloration, functional group, and
degree of oxidation of both uncapped (control) and coextruded composites before and
after UV exposures. Two separate discoloration characteristics occurred in the
discoloration of composites. For uncapped WPCs (control), chemical changes due to
photooxidation resulted in darkening followed by physical changes, including loss of
colored wood components from the surface, as well as increased roughness on the
surface, which led to lightening of WPCs. By contrast, because a hydrophobic cap
layer prevented the loss of colored components from the surface, coextruding a clear
hydrophobic HDPE cap layer over WPCs signiﬁcantly decreased the discoloration
during the weathering process. Photooxidation of wood components at the interface
accounted for the discoloration of coextruded WPCs before the failure of cap layer.
Moreover, as the cap layer absorbed a speciﬁed amount of UV light and reduced
oxygen available to interface, it decreased the photooxidation rate at the interface

compared to that at the WPCs surface.

156

5.2 Introduction

Wood ﬁber and polymers such as PVC, HDPE, and PP comprise wood-plastic
composites (WPCs). The major applications of WPCs in North America include
decking, siding, railings, fences, window and door frames, etc. These outdoor
applications expose WPCs to moisture, ﬁingi, freeze-thaw actions, and ultraviolet
light in sunlight. The presence of the hydrophobic polymer phase in WPCs improves
the durability in terms of resistance to water absorption and biological decay
compared to that of solid wood [1-4]. However, since both wood ﬁber and polymer
matrix experience photodegradation upon exposure to ultraviolet light, weathering of
WPCs results in severe discoloration and a modest loss of mechanical properties [5-
14]. A signiﬁcant feature of WPCs, aesthetic value, becomes compromised through

discoloration, making it a major concern.

Controlling the discoloration of WPCs upon weathering requires a
comprehensive understanding of weathering mechanisms. Photodegradation of WPCs,
also understood as the weathering of the major components of WPCs (wood and

polymer), results from the combined effects of light, water, oxygen, and heat.

All main polymeric components of wood, such as cellulose, hemicellulose,
lignin, and extractives, suffer from photodegradation [15]. Because visible light (400-
700 nm) carries energy less than 70 kcal/mol, it cannot provide sufficient energy to
break the major chemical bonds in the wood components [15]. Thus, the
photodegradation of wood results primarily from the ultraviolet aspect of sunlight.
The degrees of photodegradation of wood components largely depend on their ability
to absorb UV light. Because of the location of most wood chromophores in the lignin,

lignin accounts for 80-95% of light absorption of wood, making its photodegradation

157

a signiﬁcant contribution to wood discoloration [l6]. Absorption of UV light initiates
the photochemical reactions of wood surfaces leading to the formation of aromatic
and other free radicals, which causes the degradation of lignin and photo-oxidation of
cellulose and hemicelluloses, effectually causing discoloration [15, 17]. Generally,
wood in outdoor environments initially becomes yellow and brown due to the
photooxidation of lignin, and then gray due to leaching of the degraded product of
lignin [15]. This discoloration and yellowing of wood surfaces result from the loss of
methoxyl content of lignin, photo-dissociation of carbon-carbon bonds and formation
of carbonyl-based chromophoric groups. Numerous studies report that wood exposed
to artiﬁcial weathering (only exposure to UV radiation) darkens as the radiation time

increases [15-18].

Similar to the photodegradation of wood, the photodegradation of polyoleﬁns
is a radical-based oxidative process. The presence of catalyst residues, hydroperoxide
groups, and carbonyl groups introduced during polymer manufacturing, processing,
and storage accounts for this similarity [19]. All of these species absorb UV light
above 290 nm and initiate photochemical reactions. Photodegradation causes changes
in all scales of polymer dimension, including the monomer unit (oxidation), the chain
(crosslinking or chain scission), the morphology (breakdown of tie molecules and
crystal), and on the macroscopic scale [20]. Diffusion of oxygen in the polymer
controls polyoleﬁn photodegradation. Density gradients due to oxidation gradients
cause the stresses, which combined with chain scission, could initiate and propagate
cracks. Cracks on the surface result in light diffusion (a whitening effect in

appearance) and also lead to the mechanical properties loss [21].

Photodegradation of WPCs involves several factors including the wood ﬂour

content, coupling agents, manufacturing methods, and weathering conditions, among

158

 

others. Matuana and Kamdem investigated the inﬂuence of accelerated artiﬁcial
weathering on the color changes and tensile properties of PVC/wood-ﬂour composites
with various wood ﬂour contents [6]. While the composites retained original tensile
properties, they experienced severe discoloration (darkening) after weathering for
2600 hours. PVC/wood-ﬂour composites containing 15 phr wood ﬂour darken more
than those containing 30 phr and 45 phr wood ﬂour. Conversely, reports indicate the
opposite trend for weathered HDPE/wood-ﬂour composites; the composites
containing 50% wood ﬂour lighten more than the counterparts with 25% wood ﬂour
[22]. Ndiaye and coworkers studied the effects of wood ﬂour content and
compatibilizer on the durability of HDPE-based and PP-based WPCs [23]. Their
results indicated that the oxidation rate of the composites increases with wood content
and decreases with the presence of compatibilizer due to better dispersion of wood

ﬂour in polymer matrix [23].

Examining the effect of weathering conditions (UV radiation with or without
water spray) and manufacture methods (injection molding and extrusion) on the
changes in lightness and ﬂexural properties of HDPE/wood-ﬂour composites, Stark
and Matuana reported that the composites experienced more lightening and ﬂexural
properties loss when exposed to UV radiation with water spray cycle compared to UV
radiation cycle alone [ll-13]. Wood particle bleaching due to UV radiation and
removal of water-soluble extractives by water spray caused the color fading of these
composites. Not only did water spray accelerate oxidation reactions and facilitate light
penetration into the composites, but it also removed water-soluble wood extractives,
thus causing faster fading. Weathered injection-molded composites lightened more
slowly than extruded ones due to the hydrophobic HDPE-rich surface layer that

formed on the surface of injection molded composites [1 1-13].

159

Recent investigations studied the effects of natural and accelerated artiﬁcial
weathering on the appearance and chemical changes of HDPE-based and PP-based
WPCs [6-14, 24]. The lightness, total color change, and wood loss increased with
exposure time. PP-based WPCs experienced quicker photodegradation in terms of
lightness and wood loss. Deligniﬁcation and oxidation accounted for the increased
lightness of weathered composites. Muasher and Sain also investigated the
discoloration of HDPE-based WPCs exposed to natural weathering and reported that
WPCs underwent two competing redox reactions upon UV exposure: the formation of
paraquinone chromophoric structures generated by the oxidation of lignin, resulting in
yellowing, as well as the reduction of the paraquinone structures to hydroquinones,
leading to photobleaching [24]. The yellowing mechanism dominant during the ﬁrst
250 hours of exposure preceded the photobleaching mechanism that became dominant

with increased exposure time.

Current approaches of improving the weathering resistance of WPCs focus on
the bulk of WPCs, i.e., incorporation of additives into the entire product or surface
treatment of the wood ﬁber. For example, adding photostabilizers and pigments
provides protection against discoloration caused by UV radiation. Stark and Matuana
studied the effects of an ultraviolet light absorber (UVA), a colorant, and two
hindered amine light stabilizers (HALS, low molecular weight diester and secondary
amine triazine) on the photostabilization of HDPE-based WPCs composites using a
factorial design [25]. Because the UVA’s ability to absorb some UV radiation and the
colorant’s ability to block the penetration of UV radiation, as well as masking the
bleaching of wood ﬁber, UVA and colorant signiﬁcantly lowered the lightness upon
weathering. Conversely, the two HALS did not show signiﬁcant effects on lightness.

Investigation of ﬁve HALS (low, medium, and high molecular weight diester, and

160

1......“

tertiary and secondary amine triazine) and combinations of HALS and UVA on the
color changes of HDPE-based WPCs indicated that the molecular weight of diester
HALS played an important role in the color changes [24]. Low molecular weight
diester HALS shows high effectiveness in the initial stage due to their fast diffusion to
the surface but alternatively, low effectiveness with increased exposure time due to
their loss by surface evaporation [24]. Assuming the loss of color-imparting
extractives from wood ﬂour as responsible for fading, examinations studied the role
of ﬁber pretreatment on the color stability of HDPE/wood-ﬂour composites. As a
result, ﬁber pretreatment methods included removing water-soluble extractives from
wood ﬂour and coloring wood ﬂour with a water-based dye or an oil-based stain [22].
Whereas the removal of water-soluble extractives from wood ﬂour and coloring wood
ﬂour with a water-based dye did not improve the color stability of the composites
upon weathering, coloring wood ﬂour with an oil-base stain signiﬁcantly improved

the color stability [22].

Although adding stabilizer or pigment into the entire composites improves
their resistance to photodegradation, weathering primarily occurs at the surface of a
material [6-15]. Pursuing a cost-effective means to cope with weathering would not
favor adding photo stabilizers into the bulk of a material instead of the surface. The
best approach in improving durability would involve adding photo stabilizers

protection only on the surface of composite samples.

Several methods can provide a durable surface to withstand weathering.
Coating, one of the methods widely used in the wood industry, provides protection
against environmental damage such as weathering, biodegradation, and mechanical
abrasion. For protection against photodegradation, opaque pigmented coatings block

the UV/visible light but cover the aesthetics of wood. Conversely, semi-transparent

161

E

and clear coatings maintain the natural appearance of wood, but the UV/visible light
transparency of the coating results in the photodegradation of wood substrate and the

resultant failure of the coating adhesion due to the photodegraded substrate [26-29].

Investigation of the photo-discoloration of UV-curable acrylic coatings and
the underlying wood revealed that coating wood specimens with non-photostabilized
UV-curable acrylic clear ﬁlms such as aromatic urethane, aliphatic urethane, epoxy
and silicone, fails to protect wood underneath the ﬁlms against discoloration [27].
Both the coating ﬁlms and wood specimens underneath these ﬁlms experienced
severe discoloration after exposure to UVA lamps at 60°C for 12 hours.
Photodegradation of the interfacial layer led to ﬂaking and peeling of the coating.
Poor UV light resistance of the clear ﬁlms could account for the lower color stability
of these coatings, as it reduces the intensity of UV light reaching the interface layer

between wood surface and the ﬁlms [27].

In a subsequent study, investigations examined the effect of photostabilized
aliphatic or aromatic urethane-modiﬁed acrylic clear coatings on the photo-
discoloration of coated wood as a means to reduce the intensity of UV light reaching
the interfacial layer without changing the natural color of coatings and wood [29]. The
evident screening effect of ultraviolet light absorbers (UVA) such as benzophenone
and hydroxyphenyl benzotriazole resulted since their incorporation into coating ﬁlms
inhibited the photodiscoloration of wood coated with both aliphatic and aromatic
urethane-modiﬁed acrylics (19-24 h exposure). The addition of HALS failed to reduce
the discoloration of wood coated with aromatic urethane-modiﬁed acrylate containing
UVA, attributable to its inability to capture the radicals produced by the
photodiscoloration of aromatic urethane ﬁlms. In contrast, observations focused on a

synergitic effect of HALS for wood coated with aliphatic urethane-modiﬁed acrylic

162

fa ' J‘W

containing light absorbers. The addition of HALS further reduced the color changes
of coated wood due to its actions in scavenging the free radicals produced from the
photodegradation process and protecting light absorbers against the attach of free
radicals [29]. However, difﬁculty arises in coating the plastic composite at the surface
of WPCs due to the low surface energy of plastic. Coating also involves solvent
evaporation, which often causes environmental concerns. Furthermore, the drying

stage of coating requires additional time, reducing the efﬁciency of the assembly line.

Another method for providing a protective surface includes lamination. Melt
fusion or an adhesive layer allow for the lamination of a surface layer over substrate.
Nevertheless, delamination often occurs with the bending of the laminated composites
[30]. Moreover, difﬁculty arises in the usage of lamination technology to manufacture
a fully encapsulated structure in wood plastic composites rather than a planar layered

Stl'UCtUl'C.

Coextrusion, another method for providing a protective surface, can produce a
multilayered product with different properties at outer and inner layers, thus offering
different properties between surface and bulk [31, 32]. Encapsulation of WPCs with
an HDPE or PVC surface rich cap layer through coextrusion signiﬁcantly delays the
water absorption and improves ﬂexural strength compared to nonencapsulated WPCs
[33-35]. Because of water’s signiﬁcance in weathering of WPCs [1 1-13], coextruding

a hydrophobic clear cap layer onto WPCs may retard their discoloration.

Therefore, the main aim of this study examined the role coextruding a clear
HDPE cap layer onto HDPE/wood-ﬂour composites has on the photo-discoloration of
coextruded composites exposed to accelerated weathering tests in order to elucidate
the discoloration mechanisms involved in WPCs and assist in developing strategies to

reduce the weathering-induced discoloration of WPCs. Particular emphasis focused

163

11“.;

on exposing both uncapped (control) and coextruded WPCs to accelerated weathering
tests and monitoring the surface changes by color measurement, FT IR, XPS, SEM,

and UV-vis measurement.

164

5.3 Experimental
5.3.1 Materials

NOVA Chemicals supplied the neat HDPE (SCLAIR® 190) used as the cap

layer. This polymer (termed as HDPELZ), in pellet form, had a melt ﬂow index of 1.2

g/10 min and a density of 0.96 g/cm3. Another HDPE (FORTIFLEXTM 353—3511-
FLK) obtained from B.P. Solvay Polymers was used as the polymeric matrix for

uncapped composites (control) and as the core layer for coextruded composites. This

polymer (termed as HDPEOA), in ﬂake form, had a melt ﬂow index of 0.49 g/ 10 min

and a density of 0.9 g/cm3. Materials for wood ﬂour included the 0.425 mm (40 mesh)
maple from American Wood Fibers (Schoﬁeld, WI). The wood ﬂour was oven-dried

at 105 °C for ~48 h before processing to remove moisture. The lubricant TPW104 (a
blend of aliphatic carboxylic acid salts and mono and diamides) from Struktol®

Company (Stow, OH) was used to case processing and improve the surface quality of

composites.

5.3.2 Compounding and extrusion

A 20-L high intensity mixer (Papenmeier, Type TGAHK20) used for room

temperature dry blending of the HDPEOA matrix, dried wood ﬂour, and lubricant

operated for 10 minutes. The formulation of the core layer composites contained 50%

maple ﬂour, 44% HDPEOA, and 6% lubricant, based on the total weight of the

composites.

165

The coextrusion system consisted of a coextrusion die connecting a twin screw
extruder for the core layer and a single screw extruder for the cap layer, as described

in detail in previous studies [33-35].

Manufacturing of coextruded composites used core layers that contained
HDPE04/wood-ﬂour composites formed using a 32-mm conical counter rotating twin-
screw extruder (C.W. Brabender Instruments, South Hackensack, NJ) with a length-
to-diameter ratio of 13:1. The processing temperature proﬁle of the extruder
performed at 130/ 135/ 135/ 135°C from hopper to the coextrusion die and the
rotational screw maintained a constant speed at 40 rpm throughout processing. A 19
mm single-screw extruder (C.W. Brabender Instruments, South Hackensack, NJ) with
a length-to-diameter ratio of 30:1 produced the neat HDPE with MFI of 1.2 g/ 10 min
(HDPE1,2) used as cap layers. The processing temperature proﬁle of this extruder ran
at 165/ 155/ 145/ 135°C from hopper to coextrusion die. The rotational screw
maintained a speed set at 2 rpm to produce a thin cap layer (0.20 d: 0.03 mm), which

encapsulated the core layer completely (Figure 5.1).

Uncapped or noncoextruded HDPEm/wood-ﬂour composites (control) were
also produced using coextrusion die to investigate the effect of a cap layer on the
weathering of the WPCs. Manufacturing of the control samples included the twin
screw extruder with a coextrusion die using the processing conditions similar to those

mentioned above for core layers.

166

HDPE] ,2

 

(cap layer) Interface
7. i

 

 

 

 

 

 

 

HDPE0_4/wood-ﬂour composite
(core layer)

Figure 5.1. Schematic of the cross section of coextruded HDPE/wood-ﬂour

composite.

167

5.3.3 Accelerated weathering tests

An accelerated UV radiation weathering chamber simulates damaging effects
of long term outdoor exposure by exposing test samples to ultraviolet radiation,
moisture and heat. The QUV weatherometer from Q-Lab Corporation housed both
coextruded and noncoextruded control samples. Tests carried out according to the
procedures outlined the ASTM G 154 included a 12 h cycle consisting of 8h UVA-
340 radiation at 60 °C, 15 min water spray, and 3 h 45 min condensation at 50 °C.
The UVA-340 lamps give an excellent simulation of sunlight in the critical short
wavelength region from 365 nm down to 295 nm. UVA-340 radiation, water spray,
and condensation could simulate sunlight, rain, and dew, respectively. Water spray

also could introduce thermal shock and mechanical erosion.

5.3.4 Scanning electron microscopy (SEM)

SEM pictures taken using JEOL JSM-6400 SEM at 12 kv accelerating voltage
assisted monitoring the formation of cracks on the cap layer of control and coextruded
composites exposed to UV light. Moisture removal consisted of over-drying the

weathered samples at 105 °C followed by gold coating prior to observation.

5.3.5 UV-vis reﬂectance spectroscopy

A Lambda 25 UV-vis spectrophotometer (PerkinElmer) functioned as the
means in acquiring measurements of the UV-vis diffuse reﬂectance spectra of the
HDPE]; (for cap layer) and the HDPE“ (used as cap layer and polymer matrix in
both core layer of coextruded composites and uncapped WPCs). The scans were run
from 600 to 250 nm with data interval of 1 nm. The scanning speed was set at 240

nm/min.

I68

 

5.3.6 Fourier transform infrared spectroscopy

PerkinElmer Spectrum 100 with attenuated total reﬂectance (ATR) acquired
the Fourier transform infrared spectroscopy (FTIR) spectra. The surface of samples
in contact with a ZnSe crystal had a 45° angle of incidence. Scans ran at a resolution
of 4 cm'l from 4000 to 650 cm". Quantifying data used the net peak absorbance
obtained after the baseline correction. Following the approach described by Stark and
Matuana, the peak intensities of absorption bands at 1030, 1715, and 2915 cm'l were

used to calculate both wood and carbonyl indices using the following equations [8-10]:

I
Wood index = ﬂo— x100 (Equation 5.1)
12915
11715 .
Carbonyl index = ————-— x 100 (Equation 5.2)
I2915

5.3.7 X-ray photoelectron spectroscopy

A PHI 5400 ESCA spectrometer (Physical Electronics, Eden Prairie MN)
ﬁtnctioned as the instrument responsible for collecting the X-ray photoelectron

spectra. Low-resolution spectra from O to 1100 eV binding energy and high-resolution

spectra of C]S in the region from 280 to 300 eV were collected with a take-off angle
of 45°C. The degree of surface oxidation was calculated by the ratios of oxidized-to-
unoxidized carbon atoms (C1S peaks) using the following equation developed in our
previous study [6-8]:

(3 _ Coxidized _ C2 + C3 + C4
ox,*unox — C - Cl

 

(Equation 5.3)
unoxidized

169

With Carbon component CI arises from carbon atoms bonded only to carbon and
hydrogen atoms (C-C/C-H), C2 from carbon atoms bonded to a single oxygen atom,
other than a carbonyl oxygen (C-OH), C3 from carbon atoms bonded to two
noncarbonyl oxygen atoms or to a single carbonyl oxygen atom (O-C-O, C=O), and
C4 from carbon atoms, which are linked to a carbonyl and a noncarbonyl group (O-

C=O) [6].

5.3.8. Color measurements

The procedure outlined in ASTM D2244 served as the means by which to
determine the color measurements on the surfaces of coextruded and control samples.
A Minolta CR-420 Chroma Meter (Minolta Corp., Ramsey, NJ) measured the color in
L*a*b* coordinates at three locations on each sample using the Commission
lntemational d’Eclairage (CIELAB 1976 color space) three-dimension color space
system. In this system, L* axis (+L* for light, -L* for dark) represents the lightness,
a* (+a* is for red, -a* for green), and b* (+b* for yellow, and -b* for blue) the
chromaticity coordinates. At least four replicates were measured for each formulation
to obtain the average values of color. Calculations incorporated the values of lightness
and chromaticity coordinates before and after weathering tests to determine the

discoloration (AE*) of the weathered samples using the following equation [6]:

 

2 2

AB = \/AL *2 +Aa * +Ab * (Equation 5.4)

with AL*, Aa“, and Ab* as the difference of initial and ﬁnal values of L*, a*, and b*.

170

 

5.4. Results and Discussion

5.4.1 Morphology and surface characteristics of weather composites

Investigations have shown that wood ﬂour ﬁlled HDPE composites
manufactured by different processes (injection and extrusion) undergo morphological
and chemical changes in surface characteristics as they weather due to the accelerated
cyclic UV/water spray actions [8,11]. Using SEM, FT IR, and XPS techniques, Stark
and Matuana demonstrated that exposure to cyclic UV actions causes surface cracking
of the matrix and degradation of the lignin portion of the wood particles. Whereas
exposure to water spray causes wood particles to swell, which not only creates
microcracks in the matrix, but also wash away loose wood particles, loose cellulose,
and degraded lignin at the wood surface [8]. Similar techniques are used in this study
to understand the effect of coextruding a cap layer onto WPC surface on the
photodegradation of the composites. Figure 5.2 shows the micrographs of the surfaces

of both uncapped (control) and coextruded composites before and after weathering.

Prior exposure to UV/water spray cycling, coextruded WPC surface (Figure
5.2(g)) was relatively smoother than the surface of uncapped WPC (Figure S.2(a))
mainly due to the presence of a thin HDPE layer on the surface of coextruded
composites. Improper encapsulation of wood particles by the matrix could also
account for the rougher surface of uncapped WPC. As a result, cracks appeared faster
(after only 192 h) and more degradation occurred on the surface of uncapped WPC
upon accelerated weathering than the coextruded composite surface (Figure 5.2(b) to
(e)), owing to the swelling and shrinking of wood particles after absorbing and

desorbing moisture during weathering water spray cycle [8].

I71

Figure 5.2.

 

500 um

SEM (X80k) of surface of noncoextruded WPCs (left column) and
coextruded WPCs (right column) as a ﬁmction of UV radiation
exposure times: (a and f) 0 h, (b and g) 432 h, (c and h) 744 h, (d and i)
1392 h, and (e and j) 1952 h.

172

In contrast, surface cracking did not appear on the surface of coextruded
composites until 432 h of accelerated weathering (Figure 5.2(h) to 0)). The presence
of an HDPE-rich layer on coextruded WPC surface retarded the formation of cracks

in composites by delaying water sorption during weathering.

These results also imply that the transparent cap layer onto WPC did not block
all the UV radiations of UVA-340 lamps so that some ultraviolet light reached the
surface of the inner layer of coextruded composites (i.e., the surface of WPC

underneath the cap layer) during weathering experiments. Moreover, the UV-vis

spectra of the polymers used for cap (HDPE 1.2) and core (HDPEOA) illustrated in

Figure 5.3 clearly indicate that almost 85% of visible light could penetrate the
HDPE]; used as outer layer in coextruded composites. The ﬁlm absorbed only around
15-25% light in the UVA-340 radiation region (295-365 nm). Fourier transform
infrared (FT IR) spectra of HDPE matrices before and after cyclic UV/water spray
exposures corroborate this ﬁnding (Figure 5.4). Distinct absorption peaks of HDPE
are observed at around 2915, 2847, I472, 730, and 718 cm'l and Table 5.1 lists their
wavenumbers along with the assignments of corresponding ﬁrnctional groups. The
appearance of a strong peak at 1713 cm'I on the surface of HDPE (Figure 5.4(b))
clearly indicates the evidence of photodegradation of the polymer used in both core
and cap layers. Figure 5.4(a) also shows important absorption bands in FT IR spectra
associated with wood. Peaks associated with wood are clearly seen around 3354

(broad peak), 2903, 1737, 1591, and 1029 cm".

173

 

100 ’

 

 

 

,\ 80 v
83
(l)
U l
C 1
g. .
'§ 50 .. : -" HDPE1.2
2 1,1 -—HDPEO.49
:3 ..1'
h— .I
‘3
40 ‘t‘d‘?‘t‘t*t‘

 

250 300 350 400 450 500 550 600

Wavelength (nm)

Figure 5.3. UV transmission spectra of ﬁlms of HDPELZ (for cap layer) and

HDPE“ (polymer matrix for uncapped WPC and core layer of

coextruded WPCs).

174

 

 

Oi
..
, N
I I
I HOPE 1.2 In I. l
A ﬁg II I" I
b I _. _ . .I N, _ ,i k i _ ﬂ ‘ 7 ‘1 .
s l I I
0 I I
r I
b I ‘ 4": 3 i
a 3 ill- I
" HDPE 0.4 I a
c I g
0 , , 2 - 2 \' 2 , u 7 I _ ..r, _ \-
v
3 5"! I“ l
n 1
Ann- >‘\-\ E “I I l
r‘ . n r .
I \\ 8 N F P I.‘ I
mm. "our ~ N n ﬂ ., I l
4‘ .._ 1': 8 .- 1 I l 4
.- a. -. 1- , A V» . w ,.
l .f “x M ,x‘ 1“ ~ 'I’\\ 1 1:1,. \’ ”v
r, “v.1" -.-‘r"‘ 51"," .‘. v.3" vm‘k M 0‘4”” J‘ﬂvl-kxll‘ywlllvﬂ “j! \f‘jki
4000 3600 3000 2500 2W0 1500 1WD
Wavenumbors (cw-1)
I I I b
I I I ( ) l
I I II I
. 'I .
A I I 1‘ II
b . ‘ a l.
. I HDPE1.2 . g: E I
o 1 ml; " I " ,.
r . 12,-, k - \- ' I
b I __ 1 M , L Beg 3%, “WWW- l
' I l I
.. , I .
C l II 1 I
0 II II
I II I. 3
I II . I
I l V‘
'. a I
I HDPE 0.4 . r11 3 :2
I I‘ h 1- . I
I If P II "'
i ”/4,“ . ll” /‘\‘ .-

 

4000 3500 3000 2500 2000 1500 1000
Wavemlmben (cmA-1)

Figure 5.4. FTIR spectra of wood ﬂour (maple) and HDPE (a) before and (b) after

exposure to 1952 h UV.

175

Table 5.].

Assignment of absorption bands of IR spectrum.

 

 

W322???” Assignments [6-10, 36-42]
M
3354 —OH stretching in hydroxyl groups originating
mainly from cellulose
2903 CH stretching in —CH2— groups
1729-1737 C=O stretching (ester carbonyl)
1714-1717 C=O stretching (acid carbonyl)
1591 Aromatic ring stretching (lignin)
1029 C-O stretch in cellulose and deformation
in the primary alcohols of lignin
£122.12
2915 CH2 stretching of CH2, CH3 (asym.)
2847 CH2 stretching (sym.)
1713 C=O stretching (acid carbonyl)
I472 CH2 deformation (scissor vibration), crystalline
1462 CH2 deformation (scissor vibration), amorphous
730 CH2 rocking vibration, crystalline
718 CH2 rocking vibration, amorphous

 

I76

The infrared spectra of uncapped and coextruded composites are illustrated in
Figures 5.5 and 5.6, respectively. Before exposure, both uncapped (spectrum a in
Figure 5.5) and coextruded (spectrum a in Figure 5.6) composite surfaces combined
the characteristics of wood particles and HDPE matrices (Figure 5.4). However,
evidence of photodegradation to the surface of composites was apparent in the spectra
of UV-weathered samples (spectra (b) to (f) in Figures 5.5 and 5.6). The main spectral
changes are seen at around 3300-3354 cm", 1714-1738 cm'I, and 1029-1033 cm'1
frequency ranges, which are characteristics to wood components [8].

Uncapped composite surface (Figure 5.5) experienced a fast and severe wood
component loss after exposure to accelerated UV/water spray weathering than the
coextruded counterparts (Figure 5.6). Unlike coextruded composite surface, a
signiﬁcant decrease in the intensity of peaks at 3300-3354 cm'1 (OH in cellulose) and
1029-1033 cm’1 (C—O in cellulose and primary alcohols of lignin) occurred on the
surface uncapped composites even after only 192 h exposure and these peaks almost
disappeared as the weathering exposure time increased. Wood index data listed in
Table 5.2 support these ﬁndings. It is believed that water spray cycle during
weathering washed away the loose wood components (lignin decay) and the presence
of microcracks provided a channel for their fast removal on the surface of the

uncapped composites.

I77

 

onancwooc)

Figure 5.5.

If)

I to)
I (d)
I
. (c)
l — ...m ‘rﬁ 0.;
I
I 0:)
I
to)

ID 0
1- h l
h v-
. 1- h 1- I
i " if \- ._ //-¥1\- ’- _ _ H H_ -‘ ~ - ‘— . I‘ «m
I N
I '3
P
. 1 .I‘ L [I I
" I\.,, ‘ a a...“ _,r\ _2..I \2__ J \_ .. .. “4| n. I
I 1
‘ o
' . K |
1- II II
M m. \ ~— V‘“
I I.
I- K, —_. ~. ‘ -W‘ . -’ 1"\__‘ -_-_-‘ ‘\_ r“ J,‘ ﬁ‘. *4 . WK“ - _u ”‘
.__ , __ .. ”mum...” ‘—
I 1 " I
. I :2 .
.‘ ‘1 . F '.
vﬂ _ ./ ' \‘-.._ — _ﬁ_~ “ U/M__,I “*1 — — I I“ I'-_ 'I 'I‘ '7‘” \I I
I I N
In a '0 .-
8 I I',I g 8 H g '- [It
a I], v '- F: ,I 8“,.“ _J~ [ﬁr-\\x' 4 . .-
3000 2500 2000 1500 1000

Wamumben (cm“-1)

FTIR of the surface of uncapped WPCs as a function of weathering

time (a) unexposed or control, (b) 192 h, (c) 432 h, (d) 744 h, (e) 1392

h, and (f) 1952 h.

178

(f)
‘ 7 _
A , (o)
b i _
s i
o z
r (d)
b
I
n
c : (0)
g ‘ ..,.. ,. _ ,
(b)
i (a)
4000' '
Figure 5.6.

 

i i _ ' "‘—‘ ' ’
Q n '-
1 v- 3
_ ’ . ' 7 ﬁ ' / --\ ‘ .__“ \ __- HWV’mxn ' . ,
l l 2 N
8 i. " s'.
n t h # ’F,\
I I _ _ 1
t a i
r: i r
q- i ,
l | , . — .
l
'1 i
l , ‘ ' ‘-
r~ n i .
t 1‘ g f, 3 ii '
‘ - P ‘ _ W____. ’ - i ‘\ w. _ . EV
., *7 F H w ‘ ,4 .-v x. 7. M. _ a
l 1-
i a h a N l
O ; : a
8 i :2 8 q a " II?
n |,' \ F F " ‘\_ ' My,/“\_“ , .“ ‘\_- 1‘
3500 3000 2500 2000 1500 1000

Wamumbon (cm‘-1)

FTIR of the interface of coextruded WPCs as a function of weathering
time (a) unexposed or control, (b) 192 h, (c) 432 h, (d) 744 h, (e) 1392

h, and (t) 1952 h.

179

Table 5.2. Concentrations of oxidative products on uncapped and coextruded

 

 

 

 

 

 

 

WPCs.
' 0 ° 0
Exposure Wood index (/o) Carbonyl index (A)
trmes 0‘) Uncapped Coextruded Uncapped Coextruded
WPCs WPCs WPCs WPCs
O 21 30.8 6.6 8.8

432 0 36.4 16.4 11.9

1392 0 37.3 19.7 15.8

1952 O 26.7 18.8 19.0

 

180

 

Although there were cracks on the surface of coextruded composite as shown
in Figure 5.2, the presence of a thin HDPE cap layer onto composite still prevented
loose wood components from being removed by the water spray as indicated by the
intensity peak at 1029-1033 cm'1 (Figure 5.6) and percent wood index (Table 5.2),
which remained almost unchanged in coextruded composites. Two main reasons
could account for the coextruded composite surface to undergo the least wood
components loss: (i) the presence of HDPE cap layer reduced the oxygen available for
oxidation at the interface between the inner and outer layers of coextruded composites
and (ii) HDPE cap absorbed some UV energy (Figure 5.3 shows that HDPE];
absorbed 15-25% light in the UVA-340 irradiance region) to cause photodegradation.

Furthermore, composites experienced surface oxidation, irrespective of the cap
layer. This can be clearly seen by the increased carbonyl index (Table 5.2) and area
under the absorption bands in the region 1650-1750 cm", assigned to carbonyl groups
in carboxylic acid, acetyl ester, or quinine (Figures 5.5 and 5 .6). Photodegradation of
lignin leads to the formation of these carbonyl groups and thus the increase in
intensity of the carbonyl bands. Several authors have correlated the increased
integrated area or peak height of this band and carbonyl index ‘with the oxidative
photodegradation of polymer [43, 44], wood [15-17], and wood-plastic composites [9

10, 14] due to the presence of oxygenated products.

181

The results obtained from XPS spectroscopy analysis provided supporting
evidence of lignin decay and formation of oxygenated products on composite surfaces.
Table 5.3 summarizes the atomic concentrations of oxidized carbons, i.e., carbon
atoms connected to oxygen atoms (C2, C3, and C4) and unoxidized carbon (Cl)
obtained from the assignment and ratios of various peak components calculated from
the decomposed C.s peak. The C1 atomic ratio decreased after weathering, indicating
lignin decay on the surface of composites, which was accompanied by a signiﬁcant

increase in the concentration of oxidized carbon atoms (C2 + C3 + C4) or the degree

of surface oxidation (Cox/Cunox). These results suggest the presence of surface rich in

oxygenated products such as functional carbonyl groups of wood particles detected in
FTIR spectra (Figures 5.5 and 5.6). The degree of surface oxidation showed two
distinct trends. For uncapped composites, it increased with exposure time, reaching a
maximum value at 192 h, after which it decreased. This was expected because of
wood components were washed away during weathering. In contrast, the degree of
surface oxidation steadily increased in coextruded because the cap layer prevented

removal of loose wood components on the surface of the samples.

182

 

Table 5.3 Cls peak analysis at WPCs surface and coextruded WPCs interface

determined by high-resolution XPS scan.

 

 

 

 

 

 

Carbon types Exposure times (h)
0 192 432 744 1392 1952
Uncapped
WPC (control)
C1 93.6 70.7 78.0 75.4 81.8 78.7
C2 4.5 15.8 12.8 13.4 11.0 15.3
C3 2.4 7.8 4.0 7.2 3.8 4.6
C4 - 5.7 3.2 4.0 3.5 1.4
Cox/Cunox 0.08 0.42 0.26 0.33 0.22 0.27
Coextruded
composites
C1 90.5 87.1 86.9 86.6 83.8 82.4
C2 7.2 10.2 10.9 10.0 11.4 12.6
C3 2.3 2.0 2.1 2.8 3.4 3.4
C4 0 0.7 0.1 0.6 1.4 1.6
Cox/Cunox 0.1] 0.15 0.15 0.16 0.19 0.2]

 

183

5.4.2 Color analysis

Figure 5.7 illustrates the visual appearance of uncapped and coextruded
composites before and alter weathering and Table 5.4 summarizes the color

parameters measured on these samples.

Before weathering, the presence of cap layer did not affect the color of
coextruded composites since both uncapped WPCs (control) and coextruded samples
had similar visual appearance (Figure 5.7). The lightness (L*), red-green (a*) and
yellow-blue (b*) color coordinates of unexposed samples conﬁrm this observation
since both uncapped and coextruded composites had similar lightness and
chromaticity values (Table 5.4). The transparency of the thin HDPE]; ﬁlm
coextruded onto the composites could account for the appearance similarities between
the uncapped and coextruded composites before exposure to UV radiation. However,
exposure to UV radiation resulted in two distinct trends in the visual appearances of
the samples (Figure 5.7). The surface of uncapped WPC (control) lightened
continuously upon UV radiation, as indicated by a constant increase in L“ value and
signiﬁcant loss in both redness (a"‘) and yellowness (b*) (Figure 5.7 and Table 4).
Conversely, the discoloration of coextruded composites occurred in two different
stages. They ﬁrst darkened up to approximately 744 h of UV exposure and then
experienced lightening as the exposure time increased (Figure 5.7). The values of
lightness, redness, and yellowness listed in Table 5.4 conﬁrm this opposite trend.
Darkness and yellowness observed in this stage of degradation of coextruded
composites could be attributed to quinone chromophoric structures generated by the
photodegradation of lignin [15, 45, 46] whereas redness is controlled by the extractive

contents in wood [45].

184

 

Figure 5.7. Visual appearance of uncapped WPCs (left column) and coextruded
WPCs (right column) as a function of weathering exposure times (0,

192, 432, 744, 1392, 1952 h).

185

 

Table 5.4

Color parameters of control and coextruded samples before and alter

 

 

 

 

 

 

weathering.
Exposure times (h)
COIO’ 0 192 432 744 1392 1952
parameters
Uncapped
WPC
(control)
L* 59.4 70.8 81.3 83.2 84.9 88.1
a* 7.0 4.7 2.8 2.1 0.9 1.1
b* 18.2 13.1 9.3 6.5 3.4 3.9
AE - 12.6 24.0 27.0 30.1 32.3
Coextruded
composites
L“ 55.7 50.6 50.4 49.2 51.8 62.9
a* 7.2 10.0 10.5 10.9 9.8 2.0
b* 17.8 20.5 19.2 17.9 13.3 5.6
AE 6.2 6.4 7.5 6.4 15.1

 

I86

 

Color characteristics of solid wood (maple wood) and neat HDPE bars exposed to UV
radiation and water condensation were measured to understand this opposite trend.
Figure 5.8 illustrates the discoloration as well as the changes in lightness, yellowness,
and redness of these samples as a function of UV radiation time. While UV radiation
had a small inﬂuence on the discoloration of HDPE ﬁlms, it signiﬁcant impacted the
surface of solid wood (Figure 5.8), suggesting that wood components made major
contribution to the color changes in both uncapped and coextruded composites. Since
the cap layer (HDPE._2) did not fade too much and the ultraviolet light penetrates the
transparent ﬁlm (Figure 5.3), discoloration of coextruded composites could have
occurred at the interface between the wood plastic composites in the inner layer and
the HDPE]; in the outer layer.

Overall, the surface of uncapped composites (control) experienced a quick
lightening with a loss of yellowness and redness and underwent more discoloration
(AE) than coextruded counterparts (Table 5.4) due to the photodegradation of wood
components (Figure 5.8). Uncapped composites experienced quicker lightening
because water spray and condensation facilitated the removal of degraded wood
component and formation of cracks, thus accelerating fading. The loss of degraded
wood components on the surface of weathered composites could not be prevented due

to poor adhesion between the ﬁller and matrix [8].

187

 

 

 

20 100
b + WOOd ,.
—C1— HDPE1.2

15 l- 90 ..
--<>—HDPEO.4

1

AE
01
L0

10 at 80 -1
—D—HDPE1.2
—°-HDPEO.4
—0—Wood

0 1 1r 1 11 L 4 1 1 1 41 1 60 1 j1 L 14 : 1 : 1 1 1

-500 0 500 1000 1500 2000 2500 ~500 0 500 10001500 2000 2500

70‘

 

 

 

 

 

 

 

 

   
  
 

UV radiation time (hours) UV radiation time (hours)
8 30
-o—Wood 24 —o—Wood
5 ‘ —D—HDPE1.2 . —D—HDPE1.2
—<>—HDPEO.4 18 .. —<>—HDPEO.4
4 41
E 5 12 -.
2 +
6 ..
o -- MM 0 ....
' r
_2 1 1 1 1 1 1 1 1 1 1 1 _6 1 1 1 1 1 1 1 1 1 1

 

 

 

 

 

 

-500 0 5001000150020002500 -500 0 5001000150020002500

UV radiation time (hours) UV radiation time (hours)

Figure 5.8. Color parameters for solid wood (maple), HDPE0_4 and HDPE”.

188

 

However, for coextruded composites, the cap layer prevented both the removal
of both degraded wood components and wood ﬂour from the surface of weathered
samples. Consequently, color change of coextruded composites was mainly dependent
on photochemistry reactions occurring at the interface, which resulted in darkening
other than lightening. Moreover, the darkening of coextruded WPCs clariﬁed the
mechanism of fading observed in uncapped wood plastic composites (control).
Photooxidation of wood components itself would not lead to the fading of uncapped
composites since fading occurs primarily due to the loss of colored components at the
surface as well as the increase in roughness [47, 48].

Conversely, the surface of coextruded composites initially darkened and then
lightened with the failure of the cap layer. Darkening is due to the degradation of
WPCs at the interface underneath the HDPE]; cap layer since the hydrophobic HDPE
cap layer prevented removal of degraded wood components by water spray. However,
the cap layer experienced a quicker failure after 744 h weathering. Since water spray
eroded the surface, tensile and compressive stresses would be induced to the cap layer
due to the variation of temperature and relative humidity occurring during the wet
weathering cycle test [49]. Stresses damaged the cap layer integrity and accelerated
the photodegradation, resulting in the formation of cracks at the cap layer and
delamination between the cap layer and the core layer (Figure 5.2). Moreover, aﬁer
water entered the interface through cracks, it washed away the water-soluble degraded
wood component, thus causing the fading of composites underneath the cap layer of

coextruded samples.

189

 

Figure 5.9 shows the micrographs of HDPE cap layer in the “white” area and
“brown” area at the surface of coextruded samples weathered for 1392 h. Large and
deep cracks were present in “white” area (Figure 5.9(a)), implying that water could
reach the interface through these cracks and wash away degraded wood components.
It is believed that cracks at the surface of the HDPE cap layer caused light diffusion,
resulting in a whitening effect. Therefore, both the cap layer and the interface made
contributions to fading in the “white” area. By contrast, only small and narrow surface
cracks were present in the “brown” area (Figure 9(b)), which made it difficult for
water to penetrate the interface through these areas. Thus the color in these areas
appeared brown instead of white because most of chromophoric compounds generated
by photodegradation of lignin were not removed from the interface between the core

and cap layers.

190

Figure 5.9.

 

Micrograph of HDPE cap layer in “white” area (a) and “brown” area (b)

at the surface of weathered coextruded samples.

191

 

 

5.4.3. Discoloration mechanisms of WPCs

The results of this study indicate that the discoloration mechanisms of WPCs
depend on the degree of the encapsulation of the ﬁber by matrix and are related to
whether the degraded wood components remain or not on the surface of WPCs upon

UV weathering.

For uncapped WPCs, photooxidation of wood components combined with the
loss of degraded wood components due to water spray constitute the main factors for
fading, i.e., continuous lightening of WPCs during weathering. The increased
roughness of surface due to the degradation of polymer matrix could also account for

the continuous fading of uncapped WPCs.

Coextruded composites darken ﬁrst because the hydrophobic cap layer over
WPCs prevented water/moisture to penetrate the wood-plastic composites underneath
the outer layer to wash away water soluble extractives from wood and/or other
chromophoric compounds generated by photodegradation of lignin. This action
reduces the negative effect of water on weathering. in contrast, as the exposure time
increases, the cap layer cracks, water penetrates the interface washing away water

soluble extractives from wood, leading to the lightening of the interface.

These results suggest that the discoloration of WPCs upon UV weathering is a
combination of both chemical and physical changes. Chemical changes are induced
by the photooxidation of wood components, which determines the initial color change.
The composites become dark and yellow if the adhesion between ﬁbers and matrix is
strong to prevent surface erosion and removal of degraded wood components.
Otherwise, the loss of degraded wood components could occur, and if combined with

the degradation of polymer matrix which increases the roughness of WPC surface,

192

leading to in fading of the WPC surface. Therefore, whether the degraded wood
components remain on the surface of WPCs or not would determine the direction of

the long-term color change of WPCs upon UV weathering.

Based on these proposed mechanisms, several weathering-induced
discoloration phenomena reported in literature could be explained. Investigations have
shown that PVC-based WPCs darken aﬁer UV radiation [6] whereas polyoleﬁn-based
WPCs lighten [l l-l3, 22]. Owing to PVC polarity, it is believed that the interfacial
adhesion between wood ﬁber and polymer matrix is much better in PVC-based WPCs
than that in polyoleﬁn-based counterparts, making the removal of wood components
more difﬁcult on the surface of PVC-based WPCs during the weathering [6]. On the
contrary, wood components are easily washed away on the surface of polyoleﬁn-
based WPCs upon UV radiation/water spray cycle, leading to fading. It appears that
strong adhesion between wood ﬁber and polymer matrix would reduce the loss of
wood components upon weathering. Consequently, the extent of wood ﬁber
encapsulation by the polymer matrix is crucial for the discoloration of WPCs.
Generally, lower the wood ﬂour content, better the ﬁber encapsulation by the matrix,
resulting in less fading for polyoleﬁn-based WPCs [1 1-13, 22] or more darkening for

PVC-based WPCs [6].

193

5.5. Conclusions

Both noncoextruded and coextruded WPCs with a clear HDPE cap layer were
exposed to accelerated weathering cycles including UV radiation, water spray, and
condensation. The color, functional groups, and degree of oxidation of samples before
and after weathering were analyzed by chroma meter, FTIR-ATR, XPS, and UV vis.

The following conclusions can be drawn from the experimental results:

The color of noncoextruded and coextruded WPCs upon weathering changed

in different directions, revealing the discoloration mechanism of WPCs.

Overall, the surface of uncapped composites (control) experienced a quick
lightening with a loss of yellowness and redness and underwent more discoloration
(AE) than coextruded counterparts due to the photodegradation of wood components.
Uncapped composites experienced quicker lightening because water spray and
condensation facilitated the removal of degraded wood component and formation of
cracks, thus accelerating fading. The loss of degraded wood components on the
surface of weathered composites could not be prevented due to poor adhesion
between the ﬁller and matrix. This mechanism indicates the critical role of wood

ﬁber-polymer matrix interface adhesion on the discoloration of WPCs.

Conversely, the surface of coextruded composites initially darkened and then
lightened with the failure of the cap layer. Darkening is due to the degradation of
WPCs at the interface underneath the HDPE cap layer since the hydrophobic HDPE
cap layer prevented removal of degraded wood components by water spray. However,
the cap layer experienced a quicker failure after 744 h weathering. Since water spray
eroded the surface, tensile and compressive stresses would be induced to the cap layer

due to the variation of temperature and relative humidity occurring during the wet

194

weathering cycle test. Stresses damaged the cap layer integrity and accelerated the
photodegradation, resulting in the formation of cracks at the cap layer and
delamination between the cap layer and the core layer. Moreover, after water entered
the interface through cracks, it washed away the water-soluble degraded wood
component, thus causing the fading of composites underneath the cap layer of

coextruded samples.

Overall, the hydrophobic cap also prevented wood ﬁber access to
moisture/water, reducing the negative effect of water on weathering. Moreover, the
cap layer absorbed some UV light and reduced the availability of oxygen at the
interface of coextruded composites, thus decreasing the photodegradation rate. The
decreased lignin degradation rate and the oxidation rate at the interface of coextruded

WPCs were demonstrated by FTIR and XPS spectra, respectively.

5.5 Acknowledgements

The authors gratefully acknowledge the USDA-CSREES Grant-Advanced
Technology Applications to Eastern Hardwood Utilization (Grant No. 2008-34158-

19510) for the ﬁnancial support for this research work.

195

5.6 References

10.

ll.

12.

13.

J.M. Pilarski and L.M. Matuana, Durability of wood ﬂour-plastic composites
exposed to accelerated freeze-thaw cycling. Part 1. Rigid PVC matrix, Journal of
Vinyl & Additive Technology, 11(1): 1-8 (2005).

J.M. Pilarski and L.M. Matuana, Durability of wood ﬂour-plastic composites
exposed to accelerated freeze-thaw cycling. 11. High density polyethylene
matrix, Journal of Applied Polymer Science, 100(1): 35-39 (2006).

B. Dawson-Andoh, L.M. Matuana, and J. Harrison, Mold susceptibility of rigid
PVC/wood-ﬁour composites, Journal of Vinyl & Additive Technology, 10(4):
179-186 (2004).

B. Dawson-Andoh, L.M. Matuana, and J. Harrison, Susceptibility of high
density Polyethylene/wood-ﬂour composite to mold discoloration, Journal of
the Institute of Wood Science, 17(2): 1 14-1 19 (2005).

J. Markarian, Wood-plastic composites need protection, Plastics Engineering,
63(7):54-56 (2007).

L.M. Matuana and DP. Kamdem, Accelerated ultraviolet weathering of
PVC/wood-ﬂour composites, Polymer Engineering and Science, 42(8): 1657-
1666(2002)

L.M. Matuana, D.P. Kamdem, and J. Zhang, Photoaging and Stabilization of
rigid PVC/wood-ﬁber composites, Journal of Applied Polymer Science, 80(11):
1943-1950 (2001).

N.M. Stark and L.M. Matuana, Characterization of weathered wood—plastic
composite surfaces using FTIR spectroscopy, contact angle, and XPS, Polymer
Degradation and Stability, 92(10): 1883-1890 (2007).

N.M. Stark and L.M. Matuana, Surface chemistry and mechanical property
changes of wood-ﬁour/high-density-polyethylene composites aﬁer accelerated
weathering, Journal of Applied Polymer Science, 94(6): 2263-2273 (2004).

N.M. Stark and L.M. Matuana, Surface chemistry changes of weathered
HDPE/wood-ﬂour composites studied by XPS and FTIR spectroscopy, Polymer
Degradation and Stability, 86(1): 1-9 (2004).

N.M. Stark, Effect of weathering cycle and manufacturing method on
performance of wood flour and high-density polyethylene composites, Journal
of Applied Polymer Science, 100(4): 3131-3140 (2006).

N.M. Stark, L.M. Matuana, and CM. Clemons, Effect of processing method on
surface and weathering characteristics of wood-ﬁour/HDPE composites,
Journal of Applied Polymer Science, 93(3): 1021-1030 (2004).

N.M. Stark and L.M. Matuana, Inﬂuence of photostabilizers on wood ﬂour-
HDPE composites exposed to xenon-arc radiation with and without water spray,
Polymer Degradation and Stability, 91(12): 3048-3056 (2006).

196

I4.

15.

l6.

l7.

18.

I9.
20.

21.

22.

23.

24.

25.

26.

27.

J.S. Fabiyi, A.G. McDonald, M. P. Wolcott, and P. R. Grifﬁths, Wood plastic
composites weathering: visual appearance and chemical changes, Polymer
Degradation and Stability, 93(8): l405-l4l4 (2008).

W.C. Feist and D.N.-S. Hon, Chemistry of weathering and protection, US.
Department of Agriculture, Madison, (1984).

U. Miiller, M. Ratzsch, M. Schwanninger, M. Steiner, and H. bel, Yellowing
and lR-changes of spruce wood as result of UV-irradiation, Journal of
Photochemistry and Photobiology B Biolog, 69(2): 97-105 (2003).

K.K. Pandey, Study of the effect of photo-irradiation on the surface chemistry of
wood, Polymer Degradation and Stability, 90(1): 9-20 (2005).

K.K. Pandey, A note on the inﬂuence of extractives on the photo-discoloration
and photo-degradation of wood, Polymer Degradation and Stability, 87(2): 375-
379(2005)

H. Zweifel, Plastic additives handbook, Hanser Gardner, Cincinnati, (2000).

S. Girois, P. Delprat, and J. Verdu, Oxidation thickness proﬁles during
photooxidation of non-photostabilized polypropylene, Polymer Degradation
and Stability, 56(2):]69-177 (1997).

I. Yakimets, D. Lai, and M. Guigon, Effect of photo-oxidation cracks on
behaviour of thick polypropylene samples, Polymer Degradation and Stability,
86(1): 59-67 (2004).

N.M. Stark and SA. Mueller, Improving the color stability of wood-plastic
composites through ﬁber pre-treatment. Wood and Fiber Science, 40(2): 271-
278 (2008).

D. Ndiaye, E. Fanton, S. Morlat-Therias, L.Vidal, A. Tidjani, and J .L. Gardette,
Durability of wood polymer composites: Part 1. Inﬂuence of wood on the
photochemical properties. Composites Science and Technology, 68(3): 2779-
2784 (2008).

M. Muasher and M. Sain, The efﬁcacy of photostabilizers on the color change
of wood ﬁlled plastic composites, Polymer Degradation and Stability, 91(5):
1 156-1 165 (2006).

N.M. Stark and L.M. Matuana, Ultraviolet weathering of photostabilized wood-
ﬂour-ﬁlled high density polyethylene composites, Journal of Applied Polymer
Science, 90(10): 2609-2617 (2003).

B. George, E. Suttie, A. Merlin, and X. Deglise, Photodegradation and
photostabilisation of wood—the state of the art, Polymer Degradation and
Stability, 88(2): 268-274 (2005).

ST. Chang and P.L. Chou, Photodiscoloration of UV-curable acrylic coatings
and the underlying wood, Polymer Degradation and Stability, 63(3): 435-439
(1999)

197

28.

29.

30.

31.
32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

P.L. Chou, H.T. Chang, T.F. Yeh, and ST. Chang, Characterizing the
conservation effect of clear coatings on photodegradation of wood, Bioresource
Technology, 99(5): 1073-1079 (2008).

ST. Chang and P.L. Chou, Photodiscoloration inhibition of wood coated with

UV-curable acrylic clear coatings and its elucidation, Polymer Degradation and
Stability, 69(3): 355-360 (2000).

L. Long L, M.P. Wolcott, J. Zhang, and K. Englund. Flexural properties of
surface reinforced wood/plastic deck board, Polymer Engineering and Science,
47(3):281-288 (2007).

C. Rauwendaal, Understand extrusion, Hanser Gardner, Cincinnati, (1998).

H.J. Giles Jr., J.R. Wagner Jr., and EM. Mount III, Extrusion: The Deﬁnitive
Processing Guide and Handbook, William Andrew Publishing, New York,
(2004)

S. Jin and L.M. Matuana, Coextruded PVC/wood-ﬂour composites with WPC
cap layers. Journal of Vinyl & Additive T echnologt,l4(4):l97—203 (2008).

S. Jin and L.M. Matuana, Improving water resistance of wood plastic
composites Through coextrusion, SPE ANT EC Papers, 66: 19-23 (2008).

S. Jin and L.M. Matuana, Wood plastic composites coextruded with multi-
walled carbon nanotube ﬁlled rigid PVC cap layer. Polymer International, 59(5):
648-657 (2010).

M.Tsuboi, Infrared spectrum and crystal structure of cellulose. Journal of
Polymer Science, 25(109): 159-171 (1957).

K.K. Pandey KK. A study of chemical structure of soft and hardwood and wood
polymers by FTIR spectroscopy, Journal of Applied Polymer Science, 71(1):
1965-1975 (1999).

Colom X, Carrillo F, Nogues F, Garriga P. Structural analysis of photodegraded
wood by means of FTIR spectroscopy. Polymer Degradation and Stability,
80(3): 543-549 (2003).

Y. Kataoka and M. Kiguchi, Depth proﬁling of photo-induced degradation in
wood by FT -IR, Journal of Wood Science, 47(4): 325-327 (2001).

Ohkoshi M. FTIR-PAS study of light-induced changes in the surface of
acetylated or polyethylene glycol-impregnated wood. Journal of Wood Science,
48(5): 394-401 (2002).

G. Socrates, Inﬁ'ared and Raman characteristic group frequencies, John Wiley
and Sons, New York, (2004).

J .E. Mark, Physical properties of polymers handbook, Springer, New York,
(2006)

198

43.

44.

45.

46.

47.

48.

49.

SS. Femandoa, P.A. Christensena, T.A. Egerton, and J.R. White. Carbon
dioxide evolution and carbonyl group development during photodegradation of

polyethylene and polypropylene, Polymer Degradation and Stability, 92(12):
2163-2172 (2007).

L. Guadagno, C. Naddeo, V. Vittoria, C. Giovanni, and C. Cagnani, Chemical
and morphologial modiﬁcations of irradiated linear low density polyethylene
(LLDPE). Polymer Degradation and Stability, 72(1): 175-186 (2001).

L. Tolvaj and O. Faix. Artiﬁcial aging of wood monitored by DRIFT
spectroscopy and CIE L*a*b* color measurements 1: effect of UV light.
Holzforschung, 49(5): 397-404 (1995).

AJ. Ragauskas, Photoreactivity of ligninlike quinoid structures. In:
Photochemistry of lignocellusic materials, ACS symposium series, 531.
American Chemical Society, (1993).

D. Benavente, F. Martinez-Verdu, A. Bemabeu, V. Viqueira, R. Fort, M.A.
Garcia del Cura, C. Illueca, and S. Ordoﬁez. Inﬂuence of surface roughness on
color changes in building stones. Color Research & Application, 28(5): 343-351
(2009).

M. Yonehara, T. Matsui, K. Kihara, H. Isono, A. Kijima, and T. Sugibayashi,
Experimental relationships between surface roughness, glossiness and color of
chromatic colored metals. Materials Transactions, 45(4): 1027-1032 (2004).

RA. Ryntz. Plastics and coatings: durability, stabilization, testing, Hanser
Gardner, Cincinnati (2001).

199

 

CHAPTER 6

Conclusions and Future Work

6.1 Conclusions

Durability and ﬂexural properties are two key issues for application of WPCs.
Because WPCs absorb water through the surface, photodegradation of WPCs is
surface phenomenon, and surface layers of materials suffer the maximum tensile or
compressive stress when materials are bent, a durable or reinforced surface layer can
improve durability or ﬂexural properties of composites. The main objective of this
study was to improve the durability (in terms of resistance to water absorption and
weathering) and ﬂexural properties of WPCs via encapsulation of WPCs by a plastic
surface-rich cap layer through a coextrusion process. The following conclusions were

drawn from the experimental results.

1) Wood-plastic composites can be encapsulated by a neat rigid PVC or HDPE
cap layer, or a PVC-based composites (wood ﬂour or carbon nanotube ﬁlled rigid

PVC) cap layer in a coextrusion process.

2) Coextrusion of a hydrophobic polymer surface-rich cap layer over WPCs
signiﬁcantly increased moisture resistance of composites. Coextruded WPCs
absorbed less moisture and underwent a slower moisture sorption rate than
noncoextruded counterparts. While moisture resistance of coextruded WPCs was
insensitive to the cap layer thickness, moisture resistance was a strong function of the
amount of wood ﬂour in the cap layer. Moisture resistance decreased with wood ﬂour

content in the cap layer. Because the surface of coextruded samples was covered by a

200

hydrophobic polymer, the unprotected edges provided a shortcut for moisture to

diffuse into composites.

3) Coextruding a rigid PVC surface-rich cap layer on WPCs signiﬁcantly
increased their ﬂexural strength but decreased their ﬂexural modulus compared to
those of noncoextruded control. This can be attributed to greater strength but lower
stiffness of rigid PVC compared to WPCs. The changes in ﬂexural properties were
dependent on both the thickness of the cap layer and its wood ﬂour content.
Increasing the cap layer thickness did not affect the strength but decreased the
modulus. While increasing the wood ﬂour content increased the modulus but did not

inﬂuence the strength.

4) In order to obtain coextruded composites with better ﬂexural properties
than uncapped WPCs, a two-level factorial design was used to statistically evaluate
the effects of material compositions and processing conditions on these properties.
Material composition variables were the wood ﬂour content in the core layer and the
carbon nanotube (CNT) content in the cap layer of coextruded composites. The
processing condition variable was the processing temperature proﬁle for the core layer.
Regression models indicated all main effects and two powerful interaction effects
(processing temperature/wood ﬂour content and wood ﬂour content/CNT content
interactions) as statistically signiﬁcant. Factors leading to a fast fusion of the
PVC-wood ﬂour composites in the core layer, i.e. low wood ﬂour content and high
processing temperature, were effective material composition and processing condition
parameters for improving the ﬂexural properties of coextruded composites.
Reinforcing the cap layer with CNTs also produced a signiﬁcant improvement in the
ﬂexural properties of the coextruded composites, no matter the core layer composition

and the processing temperature condition.

201

5) Coextruding a transparent hydrophobic HDPE cap layer over WPCs
signiﬁcantly decreased the discoloration of WPCs upon accelerated artiﬁcial
weathering. The surface of coextruded WPCs initially darkened and then lightened
with the failure of the cap layer. Darkening is due to the degradation of WPCs at the
interface underneath the HDPE cap layer since the hydrophobic HDPE cap layer
prevented removal of degraded wood components by water spray. However, the cap
layer experienced a quicker failure after 744 h weathering. Since water spray eroded
the surface, tensile and compressive stresses would be induced to the cap layer due to
the variation of temperature and relative humidity occurring during the wet
weathering cycle test. Stresses damaged the cap layer integrity and accelerated the
photodegradation, resulting in the formation of cracks at the cap layer and
delamination between the cap layer and the core layer. Moreover, after water entered
the interface through cracks, it washed away the water-soluble degraded wood
component, thus causing the fading of composites underneath the cap layer of
coextruded samples. Overall, the hydrophobic cap also prevented wood ﬁber access to
moisture/water, reducing the negative effect of water on weathering. Moreover, the
cap layer absorbed some UV light and reduced the availability of oxygen at the
interface of coextruded composites, thus decreasing the photodegradation rate. The
decreased lignin degradation rate and the oxidation rate at the interface of coextruded

WPCs were demonstrated by FT IR and XPS spectra, respectively.

202

6.2 Future work

Coextrusion of a hydrophobic HDPE or PVC cap layer over WPCs
signiﬁcantly decreased the moisture sorption, which is critical for microbial
degradation and photodegradation of WPCs. The role of polymer-rich cap layer in the
biological resistance of WPCs needs investigation. The inﬂuences of biocide
concentration in the cap layer and fungal exposure types and duration on the decay
resistance of both control and coextruded WPCs need to be evaluated to gain an in

depth understanding of the role of encapsulation in the biological resistance of WPCs.

Although coextrusion of a transparent hydrophobic HDPE cap layer over
WPCs reduced the discoloration rate compared to the noncoextruded counterparts
upon accelerated artiﬁcial weathering, both HDPE cap layer and interface between
cap layer and core layer experienced photodegradation. Thus it is necessary to add
UV light stabilizers (UV absorbers and HALS) into cap layer to protect both the cap

layer and interface underneath the cap layer against photodegradation.

Moreover, coextruded WPCs with UV light stabilizers or biocides ﬁlled cap
layer should be compared to noncoextruded WPCs ﬁlled with UV lighter stabilizers
or biocides in the entire composites in terms of weathering or decay resistance. The
results of comparison would test whether providing a protective cap layer is more

efﬁcient to improve durability than adding preservatives to the entire composites.

In this study, we primarily focused on the properties of the cap and the core
layer but the inﬂuence of the interface between the cap and the core layer on the
properties of coextruded WPCs should not be ignored. The interface between the cap
and the core layer was already found to provide a shortcut for water sorption. The role

of the interface in ﬂexural properties should be studied.

203

 

111111111

93 03063 6520

H

s l
H
III

E “I

A “I
ll
H
I"
u

312