. Lilvufir l

1.. ..

,Jl..d:r.
Ir .a v

..0 \
.3115...”
'SJ.:I

..
.n
9 ’
z:
t .
u
. gmvwfl
1 {l-
., :19
'<
I. anxid‘
vllf a.

..
”.4 I),

.5.‘ 41’.
1: sun.“

...
.1 s.
:1" ft .
ll l.v
.2”: it
. .u;

.v i.

 

2min La AMJWWT.‘

2.2; I~ .

.4:

.9); {A t.
«I
b... 6‘
.

1":

[ THEE}?
7000

This is to certify that the
dissertation entitled

COMPOSITES FROM MALEATED POLYOLEFlN-GRAFTED
WOOD PARTICLES PRODUCED VIA REACTIVE
EXTRUSION

presented by
KARANA CARLBORN

has been accepted towards fulfillment
of the requirements for the

Doctoral degree in Forestry

 

 

(/[Ltet‘re bug; in aiua ’vz Ck

Major Piofessor’s Signature

0313 l 1 4100 g;

 

Date

MSU is an Ailirmative Action/Equal Opportunity Institution

 

LIBRARTD “ " ‘‘‘‘
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

B

07

 

C
2

CM);
01‘

1 1

I

C

C.

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/ClRC/DateDue.indd-p.1

 

COMPOSITES FROM MALEATED POLYOLEFIN-GRAFTED WOOD PARTICLES
PRODUCED VIA REACTIVE EXTRUSION

By

Karana Carlbom

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

DOCTOR OF PHILOSOPHY
Department of Forestry

2006

ABSTRACT

COMPOSITES FROM MALEATED POLYOLEFIN GRAF TED WOOD PARTICLES
PRODUCED VIA REACTIVE EXTRUSION

By
Karana Carlbom

This study examined the concept of using a reactive extrusion process to develop
a new, formaldehyde-free binding system for wood composite products. The surfaces of
wood particles were modified by grafting maleated polyethylene (MAPE) and maleated
polypropylene (MAPP) compounds through a continuous reactive extrusion process.
MAPE content was varied to study the effect of material composition on grafting
efficiency, while extruder barrel temperatures and rotational screw speeds were varied to
evaluate the effects of processing conditions on the modification of wood particles.
Polymer molecular weight effects were followed using MAPP with different molecular
weights. Efficiency of the modification was assessed using FTIR, 13C NMR and XPS
surface analysis techniques, along with a titrimetric analysis to verify the esterification
reaction between the wood particles and maleated polyolefins. Composite panels were
made from wood particles modified with MAPE and MAPP binding agents under two
different manufacturing methods. Specific contrasts of (i) base resin type, PE vs. PP, (ii)
molecular weight/maleic anhydride content in MAPP binding agents, and (iii) the
manufacturing methods (reactive extrusion vs. hot press) were investigated to determine
the effects of these factors on the physico-mechanical properties of the composites.
Finally, a response surface method using a Box-Behnken design was constructed to
statistically model and optimize the material compositions-processing conditions-

mechanical property relationships of formaldehyde-free wood composite panels.

FTIR, 13C NMR, XPS and titration data confirmed the grafting of maleated
polyolefins onto the surface of wood particles through an esterification reaction, while
the level of grafting of MAPE onto wood particles was determined to be a function of the
MAPE concentration. However, there was no significant difference found in grafting
efficiency at different extrusion processing conditions; rather all of the conditions
resulted in adequate grafting. Similarly, there was no difference in grafting efficiency
with the molecular weight of MAPP. Reactive extrusion was found to be a suitable
technique for the modification of wood particles with maleated polyolefins for all of the
material compositions and processing conditions studied. Mechanical property test
results indicated that most composite panels met or even exceeded the standard
requirements for particleboard of medium density. While extruding the particles before
panel pressing gave better internal bond (13) strength, superior bending properties were
obtained through compression molding alone. MAPP-based panels outperformed
MAPE-based panels in stiffness. Conversely, MAPE increased the 18 strength of the
panels compared to MAPP. Relationships between material compositions, processing
conditions and both flexural strength (MOR) and 18 strength of the panels were described
by linear models. Increasing any of the manufacturing variables resulted in greater MOR
and IB strength. Flexural stiffness (MOE) was described by a quadratic regression
model. Increased MOE was obtained through higher pressing times, binding agent
concentrations and/or pressing temperatures, although binding agent concentration had
less effect on MOE at higher pressing temperatures. Numerical optimization showed that
panels with desired mechanical properties could be made under a range of manufacturing

conditions.

ACKNOWLEDGEMENTS

I would first like to thank my major professor, Dr. Laurent Matuana, for his
guidance through my PhD program. He taught, me the difference between working hard
and working smart, and gave me the opportunity to develop both my research and
scientific writing skills. I am certain these skills will serve me well in my career. I
appreciate my committee members Dr. Koelling, Dr. Selke and Dr. Bix for their helpful
suggestions and advice. They spent a lot of time reading my dissertation and provided
good comments in the editing process.

My work was supported by the USDA—CSREES Grant—Advanced Technology
Applications to Eastern Hardwood Utilization, and the McIntire-Stennis Cooperative
Forestry Research Program. I am grateful to American Wood Fibers and Eastman
Chemical Company for generously donating the materials that were used in this work.

Many thanks go to my co-workers in the wood composites group at MSU. I have
been very lucky to work with such smart people and to benefit from their knowledge and
experience. I was also fortunate to work with numerous undergraduate lab assistants over
the course of this project.

Finally, I want to thank my family and friends who helped and supported me
throughout this process. I appreciate my family for always encouraging me during my
graduate studies. My friends helped me through tough days and celebrated small
victories along the way. In particular, my friends here at MSU have been invaluable in

getting through my Ph.D. program with a smile on my face.

iv

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................... viii

LIST OF FIGURES ........................................................................................................... ix

LIST OF ABBREVIATIONS .......................................................................................... xiii
CHAPTER 1

INTRODUCTION .............................................................................................................. 1

Objectives ............................................................................................................... 6

References ............................................................................................................... 7
CHAPTER 2

BACKGROUND AND LITERATURE REVIEW ........................................................... 10

Chemical Composition of Wood .......................................................................... 10

Cellulose ................................................................................................... lO

Hemicellulose ........................................................................................... 1 1

Lignin ........................................................................................................ 11

Extractives ................................................................................................. 13

Wood Anatomical Structure ................................................................................. l3

Moisture Content .................................................................................................. 17

Wood Composites ................................................................................................. 17

Medium Density Fiberboard (MDF) ......................................................... 17

Particleboard ............................................................................................. 18

Adhesives for Wood Composites .............................................................. 18

Phenolic Resins ............................................................................. 19

Amino Resins ................................................................................ 21

Wood-Plastic Composites ..................................................................................... 25

Maleated Polyolefins ............................................................................................ 26

Wood Modification with Anhydrides ................................................................... 26

Wood Modification with Maleated Polyolefins .................................................... 28

Surface Analysis Techniques ................................................................................ 31

X-Ray Photoelectron Spectroscopy (XPS) ............................................... 31

Fourier Transform Infrared (F TIR) Spectroscopy .................................... 33

Nuclear Magnetic Resonance (N MR) Spectroscopy ................................ 34

Titrimetric Analysis .................................................................................. 35

Physical and Mechanical Property Testing ........................................................... 36

Density ...................................................................................................... 36

Flexural Properties .................................................................................... 37

Internal Bond Strength .............................................................................. 37

References ............................................................................................................. 39

CHAPTER 3
COMPOSITE MATERIALS MANUFACTURED FROM WOOD PARTICLES

MODIFIED THROUGH A REACTIVE EXTRUSION PROCESS ................................ 43
Abstract ................................................................................................................. 44
Introduction ............................................................................................... 45
Experimental ............................................................................................. 48
Materials ....................................................................................... 48
Reactive Extrusion of Wood Particles .......................................... 50
Extraction of Wood Particles ........................................................ 52
Surface Characterization of Wood Particles ................................. 52
Panel Manufacturing and Mechanical Property Testing ............... 53
Statistical Analysis ........................................................................ 54
Results and Discussion ............................................................................. 55
Surface Characterization of Wood Particles ................................. 55
Mechanical Properties ................................................................... 66
Conclusions ............................................................................................... 71
References ................................................................................................. 72
CHAPTER 4
FUNCTIONALIZATION OF WOOD PARTICLES THROUGH A REACTIVE
EXTRUSION PROCESS .................................................................................................. 74
Abstract ..................................................................................................... 75
Introduction ............................................................................................... 76
Experimental ............................................................................................. 80
Materials ....................................................................................... 80
Surface Modification of Wood Particles with Maleated
Polyolefins in Reactive Extrusion ................................................. 82
Extraction of Wood Particles ........................................................ 84
Surface Characterization of Wood Particles ................................. 85
Results and Discussion ............................................................................. 89
Effect of Maleated Polyethylene (MAPE) Content ...................... 89
Effect of Extrusion Processing Conditions ................................... 99
Effect of Maleated Polypropylene (MAPP) Molecular
Weight ......................................................................................... 103
Conclusions ............................................................................................. 1 10
References ............................................................................................... l 12
CHAPTER 5

INFLUENCE OF PROCESSING CONDITIONS AND MATERIAL COMPOSITIONS
ON THE PERFORMANCE OF FORMALDEHYDE-F REE WOOD-BASED

COMPOSITES ................................................................................................................ 114
Abstract ................................................................................................... 115

Introduction ............................................................................................. 1 16

Experimental ........................................................................................... 1 20

Materials ..................................................................................... 120

Surface Characterization of Wood Particles ............................... 122

vi

Panel Manufacture ...................................................................... 124

Panel Property Testing ................................................................ 125
Statistical Analysis ...................................................................... 126
Results and Discussion ........................................................................... 127
Surface Characterization of Wood Particles ............................... 127
Physico-Mechanical Properties ................................................... 135
Density ............................................................................ 135
Effects of Processing Conditions .................................... 137
Effects of Binding Agent Compositions ......................... 140
Comparison with Standard ANSI A2081 ...................... 143
Conclusions ................................................................................. 144
References ................................................................................... 1 46
CHAPTER 6
MODELING AND OPTIMIZATION OF FORMALDEHYDE- FREE WOOD
COMPOSITES USING A BOX- BEHNKEN DESIGN ................................................. 148
Abstract ................................................................................................... 149
Introduction ............................................................................................. 1 50
Experimental ........................................................................................... 1 52
Materials ..................................................................................... 152
Experimental Design ................................................................... 152
Compounding and Panel Manufacture ........................................ 155
Property Testing .......................................................................... 156
Results and Discussion ........................................................................... 157
Mechanical Properties ................................................................. 157
Statistical Analysis of the Model ................................................ 159
Modulus of Rupture (MOR) and Internal Bond (IB)
Strength ........................................................................... 159
Modulus of Elasticity ...................................................... 162
Numerical Optimization of Mechanical Properties .................... 168
Conclusions ............................................................................................. 1 72
References ............................................................................................... 1 73
CHAPTER 7
SUMMARY OF FINDINGS .......................................................................................... 175
Future Work ............................................................................................ 180
CHAPTER 8
APPENDIX ..................................................................................................................... 183

vii

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.

Table 4.8.

Table 5.1.
Table 5.2.

Table 5.3.
Table 5.4.

LIST OF TABLES

Characteristics of Maleated Polyolefins ................................................... 49
Formulation Used for Surface Modification of Wood Particles ............... 51

High-Resolution C15 Peaks and Elemental Surface Compositions of

Wood Particles Determined by XPS ......................................................... 65
High-Resolution C1, Peaks and Elemental Surface Compositions of

Wood Particles Determined by XPS ......................................................... 69
Characteristics of Maleated Polyolefins ................................................... 82
Formulations Used for Modified Wood Particles ..................................... 84

F TIR Absorption Bands and Assignments for Unmodified Wood
Particles, Pure MAPE, Pure MAPP and Modified Wood Particles .......... 92

Elemental Surface Compositions and High-Resolution Cl, Peaks of
Wood Particles Determined by XPS. ........................................................ 96

Hydroxyl Value, Acid Value and Saponification Value Determined
by Titrimetric Analysis. ............................................................................ 97

Effect of Extruder’s Rotational Screw Speed on Surface Chemistry of
Wood Particles Modified with 20% MAPE at 160°C. ........................... 103

Effect of Extruder’s Barrel Temperature on Surface Chemistry of
Wood Particles Modified with 20% MAPE at 60 rpm. .......................... 103

Elemental Surface Compositions and High-Resolution C15 Peaks of
MAPP, Wood Particles, and Wood Particles Modified with MAPP
Determined by XPS. ............................................................................... 109

Characteristics of the Maleated Polyolefins Used as Binding Agents.... 122

Grafting Index for Peaks near 2900 cm'1 and 1740 cm'l for Unmodified
and Modified Wood Particles with Various Maleated Polyolefin
Compounds ............................................................................................. 133

High-Resolution C1, Peaks of Wood Particles Determined by XPS ...... 135

Density Data for Experimental Panels Bound with Maleated
Polyolefins .............................................................................................. 137

viii

Table 5.5.

Table 5.6.

Table 6.1.

Table 6.2.

Table 6.3.

Table 6.4.

Table A].

Table A2.

Table A3.

Table A.4.

LIST OF TABLES (CONT’D)

Effects of Processing Methods and Material Compositions on the
Mechanical Properties of Particleboard Panels Bound with Maleated
Polyolefins .............................................................................................. 1 39

Effect of Molecular Weight/Maleic Anhydride Content of MAPP on the
Mechanical Properties of Particleboard Panels Bound with Maleated

Polypropylenes ........................................................................................ 142
Box-Behnken Design Matrix in terms of Both Actual and Coded Factor
Levels Generated by Design Expert Software ........................................ 155
Standard Property Requirements for Various Grades of Particleboard of
Medium Density (640-800 kg/m3) .......................................................... 159
Analysis of Variance (ANOVA) for Response Surface Quadratic

Model ...................................................................................................... 164
Numerical Optimization Settings and Results ........................................ 171

Standard Property Requirements for Particleboard and Medium Density
Fiberboard ............................................................................................... 1 84

Mechanical Property Data for Experimental Maple Panels Bound with
MAPE ..................................................................................................... 185

Mechanical Property Data for Experimental Maple Panels Bound with
MAPP ...................................................................................................... 186

Mechanical Property Data for Experimental Panels Bound with 10.5%
MAPP, Pressed for 6 Minutes at 180°C and 3.4 MPa Pressure .............. 189

ix

Figure 2.1.
Figure 2.2.

Figure 2.3.

Figure 2.4.
Figure 2.5.
Figure 2.6.

Figure 2.7.

Figure 3.1.

Figure 3.2.

Figure 3.3.

Figure 3.4.

Figure 3.5.

Figure 4.1.

Figure 4.2.

LIST OF FIGURES

Chemical structures of (a) cellulose and (b) lignin. .................................. 12
Cross section of a ponderosa pine log showing growth rings ................... 14

SEM micrographs of (a) hardwood red oak at 100X magnification

and (b) softwood white pine at 150X ........................................................ 16
The chemistry of PF resin (a) resole and (b) novolac. .............................. 20
Chemistry of urea-formaldehyde resin ..................................................... 22
Chemistry of melamine-formaldehyde ..................................................... 24

Reaction scheme for the modification of wood particles with maleated
polyolefins ................................................................................................. 29

Modification scheme for esterification reaction between wood particles
and maleated polyolefins: (a) mono-ester and (b) di-ester formation ....... 56

FTIR spectra of unmodified wood particles (A), maleated polyethylene-
MAPE (B), MAPE-modified wood particles without extraction (C), with
24 hour-Soxhlet extraction (D), and after a second 24 hour-Soxhlet
extraction (E) in the region 4000-400 cm". .............................................. 58

FTIR spectra of unmodified wood particles (A), maleated polypropylene-
MAPP (B), MAPP-modified Wood particles without extraction (C), with
24 hour-Soxhlet extraction (D), and after a second 24 hour-Soxhlet
extraction (E) in the region 4000-400 cm‘l ............................................... 59

Solid-state '3 C NMR spectra of (A) MAPE, (B) unmodified wood
particles, and (C) wood particles modified with MAPE ........................... 63

Panel manufacture scheme for modified wood particles where R is
an ethylene or propylene repeat unit and R’ is hydrogen or a methyl

group ......................................................................................................... 67
Modification scheme for esterification reaction between wood particles

and maleated polyolefins: (a) monoester and (b) diester formation. ........ 77
Diagram of the extruder showing the three heating zones ........................ 84

Figure 4.3.

Figure 4.4.

Figure 4.5.

Figure 4.6.

Figure 4.7.

Figure 4.8.

Figure 5.1.

Figure 5.2.

Figure 5.3.

Figure 5.4.

LIST OF FIGURES (CONT’D)

F TIR spectra of unmodified wood particles (spectrum A), pure maleated
polyethylene-MAPE (spectrum B), wood particles modified with 5%
MAPE (spectrum C), 10% MAPE (spectrum D), 15% MAPE (spectrum
E), and 20% MAPE (spectrum F) after a second 24-hour Soxhlet
extraction in the region4000-400 cm". ..................................................... 91

Grafting index for FTIR absorbance bands near 2900 cm'l and 1740 cm’1
for unmodified wood particles and wood particles modified with 5-20%
MAPE ....................................................................................................... 94

Effect of processing conditions on grafting index for unmodified and
modified wood particles over the range of extruder barrel temperature

and rotational screw speed combinations: (a) band near 2900 cm'1 and

(b) band near 1740 cm". ......................................................................... 101

FTIR spectra of pure MAPP compounds with various molecular weights:
11,200 g/mol or E-43 (spectrum A), 39,000 g/mol or G-3216 (spectrum
B), 47,000 g/mol or G-3015 (spectrum C) and 52,000 g/mol or G-3003
(spectrum D) in the region 4000-400 cm". ..... ' ........................................ 105

F TIR spectra of unmodified wood particles (spectrum A), wood particles
modified with MAPP compounds of various molecular weights: 11,200
g/mol or E-43 (spectrum B), 39,000 g/mol or G-3216 (spectrum C),
47,000 g/mol or G-3015 (spectrum D), and 52,000 g/mol or G-3003
(spectrum E) in the region 4000-400 cm". ............................................. 106

Effect of molecular weight on grafting index for unmodified wood
particles and wood particles modified with 20% MAPP compounds ..... 108

Modification scheme for esterification reaction between wood
particles and maleated polyolefins: (a) monoester and (b) diester
formation ................................................................................................. 129

Panel manufacturing scheme for the two-step method showing the
bonding of pre-reacted wood particles .................................................... 129

Example FTIR spectra of unmodified wood particles (A), pure MAPEG-
2608 (B), and wood particles modified with MAPEG-2608 (C) in the
region 4000 to 400 cm" .......................................................................... 131

X-ray density profile of a sample made from unextruded MAPP
G-3003, illustrating the face and core regions of a typical sample ......... 137

xi

Figure 6.1.

Figure 6.2.

Figure 6.3.

Figure A. 1.

Figure A.2.

LIST OF FIGURES (CONT’D)

Cube graphs of the linear relationship between mechanical property
results and press temperature, pressing time, and binding agent
concentration for (a) MOR and (b) 18 strength ...................................... 162

Perturbation plot of square root of MOE against pressing temperature
(A), pressing time (B) and binding agent concentration (C) ................... 166

Interaction plots of the variation in square root of MOE as a function

of the interaction between pressing temperature and binding agent
concentration at (a) low press time (3 minutes) and (b) high press time

(9 minutes) .............................................................................................. 168

SEM images of fracture surfaces of panels manufactured from maple with
maleated polypropylene, G-3003 at 250x magnification. (a) 3%, (b)
10.5% and (c) 18% MAPP ...................................................................... 187

SEM images of fracture surfaces of panels manufactured from maple with
maleated polypropylene, G-3003 at 500X magnification. (a) 3%, (b)
10.5% and (c) 18% MAPP ...................................................................... 188

xii

AV
EPA
DRIFT
FTIR
GI
HI
IB
MAPE
MAPP
MDF
MF
MFI
MOE
MOR
MW
NMR
OSB
PB
PE
PP
PF

SV
UF
WPC
XPS

LIST OF ABBREVIATIONS

Acid Value
Environmental Protection Agency (US)
Diffuse Reflectance Fourier Transform

Fourier Transform Infrared Spectroscopy
Grafting Index
Hydroxyl Index
Internal Bond
Maleated Polyethylene
Maleated Polypropylene
Medium Density Fiberboard
Melamine Formaldehyde
Melt Flow Index
Modulus of Elasticity
Modulus of Rupture
Molecular Weight
Nuclear Magnetic Resonance (Spectroscopy)
Oriented Strandboard
Particleboard
Polyethylene
Polypropylene
Phenol Formaldehyde
Resorcinol Formaldehyde
Saponification Value
Urea Formaldehyde
Wood-Plastic Composite
X-Ray Photoelectron Spectroscopy

xiii

CHAPTER 1

INTRODUCTION

Wood-based composites are a multi-billion dollar industry in North America (1).
Manufacturing plants produce vast quantities of reconstituted wood products such as
particleboard, oriented strandboard (OSB), medium density fiberboard (MDF) and many
others. These composite products are commonly made using formaldehyde-based
adhesives, including urea-formaldehyde, melamine-formaldehyde, and phenol-
formaldehyde (1-3). In 1998, 1,780 kilotons of adhesive resin solids were used to
produce primary glued wood products (excluding the adhesive used to bond fumiture and
other secondary wood products). Of this amount, nearly 92% were formaldehyde-based
adhesives (1). Wood-based composite products find wide use in building construction,
where they are utilized as roof sheathing, wall board, floor underlayment, wood I-joists
and a number of other applications (1 -3). The furniture industry is also a major consumer
of wood composites, notably particleboard and medium density fiberboard, which are
commonly overlaid with veneer to make cabinets, shelving, tables and more (1 -3).

The use of wood composites bound with formaldehyde-based adhesives in indoor
applications is a concern because these adhesives are known to release formaldehyde
both during panel pressing and service life (1, 4-8). Plants that produce wood composites
using formaldehyde-based adhesives emit harmfiil chemicals to the environment. These
include phenol, formaldehyde, ketones, and other volatile organic compounds (1, 4-7),
which are classified as hazardous air pollutants (HAPs) by the United States

Environmental Protection Agency (EPA) (7). Wood composites bound with

formaldehyde-based adhesives also release formaldehyde over time (1, 7, 9). Although
wood composites made today emit far less formaldehyde than those made 20 years ago,
the problem has not been eliminated (1). Formaldehyde and other toxic compounds may
be present in large amounts in both indoor and outdoor air as a consequence of the use of
these adhesives (4-8).

As a result of public concern about the environment, the US EPA enacted new
emission standards for facilities that manufacture plywood and composite wood products
in September of 2004 (7). These rules affect both new and existing plants that produce at
least 10 tons of any one HAP per year, or any combination of 25 tons of HAPs per year.
These regulations may spur industry to find new ways to bind composite products
without the use of formaldehyde-based adhesives.

In recent years, there have been several studies into environmentally friendly
wood composites (1, 9). Some of the areas that have been investigated include urea-
formaldehyde adhesives with low formaldehyde-to-urea molar ratios, and the
development of natural adhesives such as tannin, lignin, soybean and cornstarch
adhesives (l, 9), phenol-formaldehyde resins modified with either lignin (10-12) or
tannin (13) and adhesive made from decayed wood (14). Additives that reduce
formaldehyde release during composite pressing and during board use have also been
developed (1, 9). Different types of fiberboard have been made without synthetic
adhesives using steam explosion (15), wood surface activation using Fenton’s reagent
(16) and wood fiber binding via lignin activation (17). A cellulose-based composite
using allyl glycidyl ether grafted polyethylene as a binder was recently reported (18).

Esterification of wood particles or other cellulosic material with anhydrides in a solvent

system has also been used as a basis for formaldehyde-free composites (19-25).
Although several approaches have been developed to reduce formaldehyde emissions,
some of them cannot be implemented in the industrial production of wood composite
products because the processes are not cost-effective, not environmentally friendly, or
lead to products with undesirable properties such as long press time, dark coloration of
the panel, low resistance to moisture, moderate strength, etc.

Control of emissions from wood composite mills and glued wood products is still
one of the major challenges facing the wood composites industry since the government is
continuously developing and implementing stringent regulations to eliminate
formaldehyde emission into the environment. A new approach to reduce and/or eliminate
formaldehyde emissions from wood composite products will be addressed in this research
project.

The technology proposed in this project is based upon the reaction between the
carboxylic acid groups of maleated polyolefins and hydroxyl groups on the surface of
wood. This reaction results in the grafting of a pendent polyolefin chain to the wood
surface through an ester linkage. Several authors have studied this reaction with
maleated compounds and wood particles (or other cellulosic materials), usually in an
attempt to evaluate maleated compounds as compatibilizers between hydrophilic wood
fibers and the hydrophobic polymer matrix in wood-plastic composites (30-35).
However, most of this previous work focused on fibers modified through a solvent-based
process where the maleated compounds were dissolved in heated xylene, toluene, N,N-
dimethylformamide (DMF) or other organic solvents, followed by the addition of

cellulosic particles (30-34). Cellulosic fibers that have been grafted with maleated

polyolefins have to be filtered and dried after modification, which makes this process
both cost-ineffective and time consuming (30-34). Solvent disposal is also an issue with
this modification process because some solvents may be harmful to workers and the
environment.

In order to produce modified wood particles that could be used to make
alternative composite panels, it is necessary to change from a solvent-based process to a
dry process. A solvent-free process that has been reported in the literature utilized a
therrnokinetic mixer to modify wood particles with maleated polypropylene (33, 35). In
that work, a heated, high-intensity mixer was used to graft maleated polypropylene onto
wood fibers. However, the esterification reaction between wood particles and maleated
polypropylene could not be confirmed through surface characterization techniques (33,
35).

An effective dry process for grafting maleated polyolefins onto the surface of
wood particles is still needed in order to produce the large volume of modified wood
particles required to manufacture and commercialize formaldehyde-free wood composite
panels. This study will investigate the use of reactive extrusion as a novel dry process for
the surface modification of wood particles needed to manufacture formaldehyde-free
wood composites.

Reactive extrusion is a technique that can be used for the chemical modification
of compounds, usually polymers that can melt during processing. Because no solvent is
involved during processing, reactive extrusion is an environmentally friendly technique
for surface modification. It also allows for control of reaction times and temperatures

through variation of rotational screw speeds and extruder barrel temperatures. Reactive

extrusion offers the advantage of intensive mixing of components in a short time,
especially in a twin-screw extruder, which ensures a more thorough reaction of
components (36). This process also has the advantage of low retention times in the
extruder, so a series of different reaction conditions can be tested in a relatively short
period of time. Examples of reactive extrusion for wood fiber surface modification were
not found in the open literature, likely because the wood component does not melt and
flow in the extruder. However, the ease of processing and control of reaction conditions
makes this technique attractive for wood fiber surface modification.

We hypothesized that maleated polyolefins would chemically react with hydroxyl
groups on the wood particle surfaces during reactive extrusion (a dry process), resulting
in wood particles with surface grafted polyolefin chains. During panel manufacturing,
the pendent polyolefin chains attached to the wood particles are expected to melt and
flow under heat and pressure in the hot press, forming entanglements. The entangled
polymer chains will then lock together after cooling, forming direct particle-particle
bonds. This process would result in a formaldehyde-free wood composite panel

manufactured without any additional adhesive.

Objectives

The main goal of this work is to study the concept of using a reactive extrusion
process as a means of developing a new, formaldehyde-free binding system for wood
composite products. The following specific objectives must be accomplished to achieve

the main goal of this project:

1. Evaluate the effects of material composition (binding agent types and content)
and extrusion processing conditions (temperature profile and rotational screw
speed) on the level of grafting (surface properties) of wood particles after the

reactive extrusion process;

2. Characterize the surface of unmodified and modified wood particles in terms of

chemical compositions (both elemental and functional groups);

3. Manufacture composites and evaluate their physico-mechanical properties;

4. Establish the relationships between material composition, processing conditions,
and physico-mechanical properties in order to identify the major factors that
govern the performance of formaldehyde-free wood composite products

manufactured through a reactive extrusion process.

References

10.

11.

12.

13.

Sellers, T., Jr., "Wood adhesive innovations and applications in North America,"
Forest Products Journal, 51 (6): 12-22 (2001).

Maloney, T.M., "The family of wood composite materials," Forest Products
Journal, 46 (2): 19-26 (1996).

Guss, L.M., "Engineered wood products: the future is bright," Forest Products
Journal, 45 (7/8): 17-24 (1995).

Barry, AD. and Comeau, D., "Volatile organic chemicals emissions from OSB as
a function of processing parameters," Holzforschung, 53 (4): 441-446 (1999).

Barry, A., Comeau, D., and Lovell, R., "Press volatile organic compound
emissions as a function of wood particleboard processing parameters," Forest
Products Journal, 50 (10): 35-42 (2000).

Barry, A., Lepine, R., Lovell, R., and Raymond, S., "Response surface
methodology study of VOCs in plywood press emissions," Forest Products
Journal, 51 (1): 65-73 (2001).

US-EPA, "National emission standards for hazardous air pollutants: plywood and
composite wood products," Federal Register, 69 (146): 45943-46046 (2004).

Maloney, T.M., Modern Particleboard and Dry-Process F iberboard
Manufacturing, Updated Ed., Miller Freeman: San Francisco (1993).

Pizzi, A., Wood Adhesives : Chemistry and Technology. Marcel Dekker: New
York (1983).

Matuana, L.M., Riedl, B., and Barry, A.O., "Kinetic characterization by DTA of

lignosulfate-based phenol-formaldehyde resins," European Polymer Journal, 29
(4): 483-90 ( 1993).

Kazayawoko, J.S.M., Riedl, B., Poliquin, J ., Barry, A.O., and Matuana, L.M., "A
lignin-phenol-formaldehyde binder for particleboard. Part 1. Thermal
characteristics," Holzforschung, 46 (3): 257-62 (1992).

Sellers, T., Jr., McGinnis, G.D., Ruffin, T.M., and Janiga, E.R., "Lignin-modified
phenol-formaldehyde resin development for fiberboard," Forest Products
Journal, 54 (9): 45-51 (2004).

Sellers, T., Jr. and Miller, G.D., Jr., "Laboratory manufacture of high moisture
southern pine strandboard bonded with three tannin adhesive types," Forest
products journal, 54 (12): 296-301 (2004).

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

Li, K. and Geng, X., "Formaldehyde-free wood adhesives from decayed wood,"
Macromolecular Rapid Communications, 26 (7): 529-532 (2005).

Velasquez, J .A., Ferrando, F., Farriol, X., and Salvado, J ., "Binderless fiberboard
from steam exploded Miscanthus Sinensis," Wood Science and Technolog, 37
(3-4): 269-278 (2003).

Widsten, P., Qvintus-Leino, P., Tuominen, S., and Laine, J .E., "Manufacture of
fiberboard from wood fibers activated with F enton's reagent (H202/FeSO4),"
Holzforschung, 57 (4): 447-452 (2003).

Yelle, D., Goodell, B., Gardner, D.J., Amirbahman, A., Winistorfer, P., and
Shaler, S., "Bonding of wood fiber composites using a synthetic chelator-lignin
activation system," Forest Products Journal, 54 (4): 73-78 (2004).

Casarano, R., Matos, J .R., Fantini, M.C.A., and Petri, D.F.S., "Composites of allyl
glycidyl ether modified polyethylene and cellulose," Polymer, 46 (10): 3289-3299
(2005)

Matsuda, H., "Preparation and utilization of esterified woods bearing carboxyl
groups," Wood Science and Technology, 21 (1): 75-88 (1987).

Matsuda, H., Ueda, M., and Mori, H., "Preparation and crosslinking of
oligoesterified woods based on maleic anhydride and allyl glycidyl ether," Wood
Science and Technology, 22 (1): 21-32 (1988).

Matsuda, H., "Thermal plasticization of lignocellulosics for composites," in
Emerging Technologies for Materials and Chemicals from Biomass R.M. Rowell,
T.P. Schultz, and R. Narayan, Editors. American Chemical Society: Washington,
DC. p. 98-114 (1992).

Clemons, C., Young, RA, and Rowell, R.M., "Moisture sorption properties of
composite boards from esterified aspen fiber," Wood and Fiber Science, 24 (3):
353-63 (1992).

Timar, M.C., Mihai, M.D., Maher, K., and Irle, M., "Preparation of wood with
thermoplastic properties. Part 1. Classical synthesis," Holzforschung, 54 (1): 71-
76 (2000).

Timar, M.C., Maher, K., Irle, M., and Mihai, M.D., "Preparation of wood with
thermoplastic properties. Part 2. Simplified technologies," Holzforschung, 54 (1):
77-82 (2000).

Timar, M.C., Maher, K., Irle, M., and Mihai, M.D., "Thermal forming of
chemically modified wood to make high-performance plastic-like wood
composites," Holzforschung, 58 (5): 519-528 (2004).

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

Hon, D.N.S. and Xing, L.M., "Thermoplasticization of wood. Esterification," in
Viscoelasticity of Biomaterials, W.G. Glasser and H. Hatakeyama, Editors.
American Chemical Society: Washington DC. p. 118-132 (1992).

Marcovich, N.E., Reboredo, M.M., and Arangguren, M.I., "Sawdust modification.
Maleic anhydride chemical treatment," Holz als Roh- und Werkstojf 54 (3): 189-
193 (1996).

Marcovich, N.E., Aranguren, M.I., and Reboredo, M.M., "Modified woodflour as
thermoset fillers Part I. Effect of the chemical modification and percentage of
filler on the mechanical properties," Polymer, 42 (2): 815-825 (2001).

Marcovich, N.E., Reboredo, M.M., and Aranguren, M.I., "Modified wood flour as
thermoset fillers. 11. Thermal degradation of wood flours and composites,"
T hermochimica Acta, 372 (1-2): 45-57 (2001).

Felix, J .M. and Gatenholm, P., "The nature of adhesion in composites of modified
cellulose fibers and polypropylene," Journal of Applied Polymer Science, 42 (3):
609-20 (1991).

Matuana, L.M., Balatinecz, J .J ., Sodhi, R.N.S., and Park, C.B., "Surface
characterization of esterified cellulosic fibers by XPS and FTIR Spectroscopy,"
Wood Science and Technology, 35 (3): 191-201 (2001).

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

Kazayawoko, M., Balatinecz, J .J ., and Woodhams, R.T., "Diffuse reflectance
Fourier transform infrared spectra of wood fibers treated with maleated
polypropylenes," Journal of Applied Polymer Science, 66 (6): 1163-1173 (1997).

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

Kazayawoko, M., Balatinecz, J .l ., and Sodhi, R.N.S., "X-ray photoelectron
spectroscopy of maleated polypropylene-treated wood fibers in a high-intensity
therrnokinetic mixer," Wood Science and Technology, 33 (5): 359-372 (1999).

Cheremisinoff, N.P., Guidebook to Extrusion Technology. Prentice Hall:
Englewood Cliffs, NJ. (1993).

CHAPTER 2

BACKGROUND AND LITERATURE REVIEW

Chemical Composition of Wood

Wood is a complex, multi-component material composed of several compounds.
Wood is primarily composed of cellulose, hemicellulose and lignin, with minor
components extractives and ash. Although amounts of these components vary between
species, wood composition is approximately 40-45% cellulose, 15-30% hemicellulose,

20-30%1ignin, 1-10% extractives and 0-1% ash (1-4).

Cellulose

Cellulose is the primary component of wood. Carbon, hydrogen and oxygen are
arranged into sugar molecules during photosynthesis in the living tree. These B-linked
glucose units are further linked together to form long, linear chains. Cellulose chains are
generally oriented in one direction, and give strength to the cell wall. The orientation of
cellulose is largely due to the presence of a large number of hydroxyl groups. This
allows the chains of cellulose to interact and hydrogen bond with adjacent chains,
forming crystalline regions within the wood. However, not all of the cellulose is ordered
into crystalline domains. Some portion of cellulose is amorphous, and in these regions
the hydroxyl groups are free to form hydrogen bonds with water molecules. This
accounts for the hygroscopic nature of wood, as water is held in the amorphous portions
of cellulose. Water cannot easily enter the crystalline portions of wood as the cellulose

chains there are tightly packed and held together by hydrogen bonding across their

10

hydroxyl groups. There are essentially no free hydroxyl groups available to hold water in
the crystalline regions of wood. The amorphous region of cellulose is also the reactive
site for wood modification as the hydroxyl groups are accessible for reaction with other

compounds (1-4). Cellulose structure is illustrated in Figure 2.1a.

Hemicellulose

Hemicellulose is structurally similar to cellulose, as it is also composed of sugar
molecules. However, hemicellulose is formed from various 5 and 6 carbon sugars,
connected in short, often branched chains. Hydroxyl groups are also present in the
hemicellulose, accounting for some of the hygroscopic nature of wood. Because of the
short chain structure, most of the hemicellulose is soluble in water, and can be removed

by water extraction (1-4).

Lignin

Like cellulose and hemicellulose, lignin is formed from carbon, hydrogen and
oxygen. However, instead of taking the form of a polysaccharide, it is a phenolic
compound which exists in a variety of forms. Phenylpropane units form the basic
structure, linked together and branched in various ways. The structure of lignin is
illustrated in Figure 2.1b. Lignin is primarily concentrated in the outer portions of wood
cells and between the wood cells, where it fimctions as a binding agent to hold cells
together. It is a brittle, stiff material that strengthens the cell. There are fewer hydroxyl

groups in lignin than cellulose, which makes lignin less reactive (1-4).

11

3
/°\

:13
O
l:\
O
M

3
0—0

5°

:1:
o’ \a;_e

230:1:
O
m

‘2
o
:1:/\
6
:1:

0—0

 

nocrr2

 

Figure 2.1. Chemical structures of (a) cellulose and (b) lignin (3)

12

Extractives

Extractives are a class of organic compounds that includes resins, carbohydrates,
waxes, tannins, fats, oils, acid, etc. They are called extractives because they are not
strongly bound to the wood structure and thus can be easily removed or extracted through
processes such as steam distillation and solvent extraction. Extractives are largely
responsible for the smell, color, density, flammability and fungal resistance of wood (1-

4).

Wood Anatomical Structure

Growth rings are one of the most distinctive features of wood. They are
noticeable on a cross section of wood or on the stump remaining when the tree is cut.
Figure 2.2 shows an example of a tree cross section with clearly defined growth rings.
The age of a tree can be determined by counting the rings, as each ring represents one
year of growth. The light-colored portion of the ring is called earlywood and is
composed of large, conductive cells that form early in the growing season. In contrast,
the dark portion of the ring is latewood, which is composed of cells with thicker walls
and more strength. The thickness of the rings and the proportion of earlywood to

latewood varies between species, and is also affected by growing conditions (3, 4).

13

 

Figure 2.2. Cross section of a ponderosa pine log showing growth rings (4)

 

Opt
0f
V65

Ba

31

The cells of wood are long and narrow with hollow centers called lumens.
Openings called pits connect the cells along their length (3, 4). There are two main types
of cells within wood: (i) fibers and (ii) vessels. Fibers give the wood strength, while
vessels conduct liquids. Both the size and quantity of these cells differ within a tree.
Based on these differences, wood is categorized as softwood or hardwood (3, 4).

Sofiwoods, or coniferous trees, are characterized by a porous structure formed by
a large quantity of fiber tracheids. There are no vessels present, although some softwood
has resin canals that transport materials in the tree (3, 4). Softwood fibers are rectangular
and very long, approximately 3-8 mm, and are oriented in nearly straight rows (3, 4).
The fiber tracheids lend strength to the tree, while also conducting liquids in the vertical
direction. Groups of cells called rays are also present to move sap horizontally in the
living tree (3, 4). Softwood structure is illustrated in Figure 2.3.

Hardwoods, or deciduous trees, contain both fibers and vessels, randomly
oriented within the wood. An example of hardwood structure is illustrated in Figure 2.3.
Hardwood fibers are circular and shorter in length than those in softwood (3, 4). A
typical hardwood fiber averages only 1 mm in length (3, 4). Vessels form vertical canals
for the transport of materials within the tree. Hardwoods also have rays to move

materials horizontally through the tree (3, 4).

15

 

Figure 2.3. SEM micrographs of (a) hardwood red oak at 100X magnification and (b)
softwood white pine at 150x (3)

 

 

 

 

 

 

 

 

Mo

 

 

Moisture Content

Wood is a hygroscopic material (1-4). As a result, wood moisture content varies
depending on the moisture content and temperature of its surroundings. The fiber
saturation point is defined as the point where the lumens do not contain any free water
but the cell wall is fully saturated (3). Below this point, moisture content of wood plays a
large role in the dimensional stability, as wood shrinks and swells in response to moisture
changes (3). Shrinking and swelling of wood elements can cause problems when wood is
used in composites, as these changes weaken glue bonds and reduce interfacial adhesion

between fibers and matrix components.

Wood Composites

Glued wood composites can be made from wood of varying geometry, including
products based on lumber, sheets of veneer, strips of wood, flakes, particles, fibers and
even wood flour (3-7). In the case of this research, wood particles were used to produce
formaldehyde-free composites via reactive extrusion. The extrusion process limits the
size of the wood material that can be used, as larger fibers or particles would be broken
down by the shearing action of the screws in the twin-screw extruder. Glued wood
composites made from wood particles or fibers would be most similar to the panels

produced via reactive extrusion, thus these will be described in detail.
Medium Density F iberboard (MDF)

MDF is a panel product composed of wood fibers and a thermoset adhesive.

Urea-formaldehyde (UP) is the most common adhesive used in MDF, although some

17

isocyanate-p-MDI is used for certain applications. Panels are hot pressed to cure the
resin and consolidate the fibers to the medium density range, between 500 and 800 kg/m3.

MDF panels are often covered with veneer and used in cabinetry and moldings (6, 8).

Particleboard

Particleboard is a panel product made from discrete particles of wood and UP
adhesive. The wood and adhesive mixture is formed into a mat and then pressed in a
heated press to consolidate the panel. Particleboard is used in some structural
applications, such as flooring and wallboard, although it finds a majority of its use in
furniture applications such as shelving, in which it is covered with a layer of veneer or

laminate (6, 8).

Adhesives for Wood Composites

Two categories of adhesives dominate in the production of glued wood composite
products. These are phenolic resins and amino resins, which combined made up over
90% of the adhesives used in wood composites in 1998 (7). Phenolic resins include
phenol-formaldehyde (PF) and resorcinol-formaldehyde (RF), whereas amino resins
include urea-formaldehyde (UP) and melamine-formaldehyde (MF) (3, 7, 9). Polymeric
diphenyl methylene diisocyanate (PMDI) is also used for some specialty applications (3,

7). Phenolic and amino resins are discussed in detail.

18

Phenolic Resins

Phenol-formaldehyde (PF) resins are made by the reaction of phenol with
formaldehyde. There are two basic types of PF resin, resole and novolac (3, 9). The
structures of these resins are shown in Figure 2.4.

Resole PF is made through the reaction of phenol with excess formaldehyde in the
presence of an alkaline catalyst. This results in a resin with a branched structure. The
presence of reactive methylol groups allows this type of PF to self-cure with the
application of heat, and thus it does not need an added catalyst (3). Resoles are dark
yellow to dark brown in color.

Novolac PF resins are formed from the reaction of excess phenol with
formaldehyde under acidic conditions (3). This produces a resin with a linear structure.
Novolacs cannot self-cure because there are no residual methylol groups in the structure.
As a result, a curing agent must be added to these resins in order to form crosslinks.
These resins are lighter in color than the resole PF and also have good moisture
resistance.

Resorcinol-formaldehyde (RF) is another type of phenolic resin used in the wood
industry. This adhesive is produced by the reaction of resorcinol with formaldehyde. RF
resins have two reactive hydroxyl groups in their structure, which allows them to cure
rapidly at room temperature with the addition of a catalyst (3). RF is expensive to
produce, but is still desirable for high moisture resistance of products made with this

adhesive (3).

19

OH

Phenol Formaldehyde Monomethylolphenol

 

:—n—:
5
0-2
$-61
2

Hon/— Methylene bridge

OH

110

""" alk

'40 _________ ©—E—Ofl__ Z—OH + HOH
O" P: F<l

Heat

©Resol©e Molecular grth

and cross-linking

¢ Edi.“

+ XHOH

0H

   

OH OH

Ito" acid
(b) H + + HOH
P: F > 1
Heat and
Novolac formaldehyde
donor

Molecular grth
0H 1" OH I“ (RH I"! OH I. OH I" PH
Oi at: cou *

Figure 2.4. The chemistry of PF resin (a) resole and (b) novolac (3)

20

Am

 

ad’:

8111

for

C0-

bi

 

 

 

 

Amino Resins

Urea-formaldehyde (UF) resins are the least expensive of the major wood
adhesives. They are formed from the polymerization of excess formaldehyde with urea,
and react as the four N-H bonds on the urea add across the carbonyl group on the
formaldehyde (Figure 2.5). This produces methylol groups, which are suspended in
water at about 65% resin solids when sold. The methylol groups crosslink by
condensation when an acid catalyst is added, and after final curing the resin is hard,
brittle and insoluble. However, cured UP is sensitive to moisture, which causes the
resin to break down and release formaldehyde (3, 9). Formaldehyde release from UF
bonded wood products is also a concern, especially in light of the new EPA regulations

on formaldehyde emissions from wood composite manufacturing facilities (10).

21

iii-P * lr—riiif"

Urea Formaldehyde Monomethylolurea

H

Ill—:1 i'iiom'f'iiJN-c-r't {mu

0 ll H
m-itriri7 *"°"
9’“ Molecular growth
H- H and cross-linking
+30
l-H’
H-C-H
iiiii li.i.i..
......... “-c- - - - - - -
"i Til it. 3 M43
T-H “‘30 + XflOl-l
02C 0 H at?" I! H
ii"- -12-- 3.2-2- -2-.-,u-.-§-.~.-i...
a ti 3 ti 1 H a
ll- -ll H | :31!
0-3 fl-N '
i-0 ,3-
Sin eel.
6-n 0-3

Figure 2.5. Chemistry of urea-formaldehyde resin (3)

22

Melamine-formaldehyde (MF) resins are also widely used in wood composites.
These resins are produced through the reaction of melamine with formaldehyde as
illustrated in Figure 2.6. This produces a resin that is somewhat water resistant because
of the low solubility of melamine in cold water. MF glues can be used for interior and
exterior applications, but their high cost is a large drawback. MF resins are more
expensive than PF, and offer less durability for an increased price. Often, MF and UP
resins are blended to produce a resin with lower cost than MF with improved properties

compared to UP alone (3, 9).

23

 

 

Ii

 

 

H
l

I

" i

/

h/ \N r/ N
Ii! 3 I I" T I! H J H II! Polymerizes and
H—N— C— N—H + (II=O —~ H—N— —N—C—OH-—-° cross-links like
I .
\/ H \/ H urea resrn
Melamine Formaldehyde Monomethylol melamine

Figure 2.6. Chemistry of melamine-formaldehyde resin (3)

24

Wood-Plastic Composites

Wood plastic composites (WPCs) are a versatile family of composite materials
made from wood fibers, particles, or flour mixed with a thermoplastic polymer and
formed into a variety of shapes (11, 12). WPCs are used in a variety of applications such
as decking, fencing, docks, automotive applications, playground equipment, etc (12).
The decking market has been the largest volume application for WPCs, consuming over
50% of the total WPC produced in 2002 (12). When well—made, WPCs have several
advantages over wood or plastic alone such as resistance to fungal decay, high impact
strength, low incidence of cracking, etc (1 l, 12). WPCs are usually manufactured by
extrusion, where wood and plastic are melted together and compounded into a composite
material (12). The composite can be extruded into lengths of material that resemble
lumber, or into profiles with hollow interiors or various layers. More complex shapes can
be produced through injection molding, in which the composite is injected into a mold
while still melted. The compounded mixture cools in the mold and is removed once
solid.

Despite the advantages offered by WPCs, one significant drawback to their use is
the lack of compatibility between wood, which is hydrophilic due to the large content of
cellulose, and the hydrophobic plastic matrix (13-19). This incompatibility results in
reduced strength properties and lower quality products. In order to overcome this
drawback, coupling agents have been developed to compatibilize the wood and polymer
components (13-19). The most successful examples of WPC coupling agents for
polyolefin-based WPCs are maleated polyolefins, which bond with the wood component,

while diffirsing into the plastic matrix.

25

Maleated Polyolefins

Maleated polyolefins are usually polypropylene or polyethylene polymer that has
been grafted with a small amount of maleic anhydride. These polymers are mainly used
as coupling agents to promote interfacial adhesion in wood polymer composites between
the non-polar plastic and polar wood components (13-19). The carboxylic acid portion of
the maleated compound reacts with hydroxyl groups on the wood surface to form an ester
linkage (13, 14, 17). The polymer chain is then free to diffuse into the polymer matrix,
forming a physical bond upon cooling. This coupling mechanism is an effective method
to promote compatibility between the phases in wood composites (13, 19-21). A small
amount of maleated polyolefins (1-5%) is generally used to promote adhesion between

the phases (13, 17, 19, 20).

Wood Modification with Anhydrides

Another method of compatibilizing wood and plastic in composite materials is to
chemically modify the wood component. Chemical modification of wood with anhydride
compounds has been studied extensively for several decades (22-32). Wood reacts with
various anhydrides to form ester groups on the wood surface. Esterification has been

shown to plasticize the wood, making it more versatile for wood composite applications.

Several investigators have studied wood modification with anhydrides in order to
determine thermal properties, moisture resistance, compatibility with polymeric matrices,
etc. (22-32). A lower softening temperature, compared to unmodified wood, has been
documented in esterified wood after reaction with various anhydrides (22, 27-29).

Esterified wood has also been found to resist moisture (22-28, 30-32) and to be more

26

 

C01

C01

for

C01

1hr

ch

\l'

Ci

Eli

 

 

compatible than unmodified wood with polyester matrices in the manufacture of
composites (22-28, 30-32). A second reaction of the esterified wood with epoxides
formed oligoesters, which further thermoplasticized the wood and allowed for better

control of structure and resulting properties of the modified wood (22-24, 26-28).

While the thermoplastic nature of esterified and oligoesterified wood modified
through these methods has been documented, few authors have exploited this
characteristic to make wood composites without added adhesive. Matsuda and co-
workers used three dicarboxylic acid anhydrides (maleic, phthalic and succinic
anhydride) to esterify wood particles, which were then molded into sheets through
compression molding (22, 24). Clemons and co-workers used similar methods to modify
wood fibers with anhydrides to prepare fiberboards (25). In both studies, chemically
modified wood was able to partially melt when heated and bond without additional
adhesive. Formulations containing succinic anhydride were found to have the most
thermoplastic character in both studies (22, 24, 25). Further work by Matsuda et a1. (22-
24) showed enhanced thermoplasticization of wood through the grafting of various types
of epoxides onto the already esterified wood surface. The oligoesterified wood produced
from these reactions was even more thermoplastic-like in structure than the esterified
wood produced through reaction with the anhydrides, and could be easily molded. Using
various types of epoxides allowed for different chemical structures in the
thermoplasticized wood, and could also be used to form crosslinked wood composites
which resembled plastic more than wood (22-24). Recently, Timar and co-workers
demonstrated the use of maleic anhydride to esterify wood particles, followed by

oligoesterification with glycidyl methacrylate and additional maleic anhydride as a

27

process to form thermoplastic-like wood particles (26-28). Panels formed through the
compression molding of modified wood particles displayed mechanical properties that
met or exceeded standard requirements for bending strength and internal bond strength of
particleboard or fiberboard (28). These wood composites were also found to be resistant

to fungal decay.

Although esterifying and oligoesterifying the wood surface has been shown to be
an effective method for manufacturing wood composite materials without additional
adhesive, most methods require harmful organic solvents such as xylene, dimethyl
sulfoxide, or N,N-dimethylformamide as part of the modification process (22, 24-32).
Additionally, the oligoesterification reactions entail a two-step process (or more
depending on the precise control of chemical structure desired) to functionalize the wood
surface, which is both time consuming and complicated (22-24, 26-28). Removal of
solvents and further drying of the wood particles is required before compression molding

into panels (22. 24, 25, 28).

Modification of Wood with Maleated Polyolefins

Maleated polyolefins also react with wood particles through an esterification
reaction (Figure 2.7). This reaction occurs between the maleated groups of the
polyolefins and hydroxyl groups on the wood surface, forming a monoester (Figure 2.7.a)
or a diester (Figure 2.7.b). This results in the grafting of a pendent polyolefin chain to the

wood surface through the ester linkage.

28

////

 

 

 

 

ii
/ lFO‘C_ Cc? Monoester

Polyolefin ; OH CH M
0 chain [— /
a) / HO’CE
l
/ —0H %CH2 1 / O
/ t 0 CIH or
; _0H % \/W \ C? Polyolefin
I .
O b) / O—~C—CH2 chain
Wood Maleated /
particle polyolefin / o— c —C
II H
/ o
Modified wood Diester
particles

Figure 2.7. Reaction scheme for the modification of wood particles with maleated
polyolefins

29

Several authors have studied this reaction with maleated compounds and wood
particles (or other cellulosic materials), usually in an attempt to evaluate maleated
compounds as compatibilizers between hydrophilic wood fibers and the hydrophobic
polymer matrix in wood-plastic composites (13-18). However, most of this previous
work focused on fibers modified through a solvent-based process where the maleated
compounds were dissolved in heated xylene, toluene, N,N-dimethylformamide (DMF) or
other organic solvents, followed by the addition of cellulosic particles (13-17). As with
wood particles esterified or oligoesterified with other anhydride compounds, cellulosic
fibers that have been grafted with maleated polyolefins have to be filtered and dried after
modification, which makes this process both cost-ineffective and time consuming (13-
17). Solvent disposal is also an issue with this modification process because some
solvents may be harmful to workers and the environment.

In order to produce modified wood particles that could be used to make
alternative composite panels, it is necessary to change from a solvent-based process to a
dry process. A solvent-free process that has been reported in the literature utilized a
thermokinetic mixer to modify wood particles with maleated polypropylene (16, 18). In
that work, a heated, high-intensity mixer was used to graft maleated polypropylene onto
wood fibers. However, the esterification reaction between wood particles and maleated
polypropylene could not be confirmed through surface characterization techniques (16,

18).

30

 

Surface Analysis Techniques

Several techniques are employed in the surface analysis of wood-based materials.
These techniques include X-ray photoelectron spectroscopy (XPS), Fourier transform
infrared (FTIR) spectroscopy, and nuclear magnetic resonance (NMR) spectroscopy.
These techniques are frequently used together to determine chemical composition,
structure and information about reactions in materials. An additional chemical
characterization technique for wood-based materials is titration analysis, which is used to
determine bulk chemical changes through acid-base chemistry. Contact angle
measurements are often used to determine wettability and surface tension of wood-based
materials. Inverse gas chromatography can be also used to determine thermodynamic
properties of modified wood particles. Techniques used to analyze chemical changes
resulting from modification of wood particles in reactive extrusion with maleated

polyolefins in this work will be discussed in more detail.

X-Ray Photoelectron Spectroscopy (XPS)

XPS is a technique that allows chemical elemental composition of samples to be
analyzed in a quantitative manner. A beam of x-ray photons is aimed at the sample,
which interacts with the surface and causes electrons to be emitted. The binding energy
(Eb) of the emitted electrons is measured by the detector. Since each element has a
characteristic Eb, XPS can be used to determine the elemental composition of a sample.
This technique is surface sensitive, with a maximum sampling depth of approximately 10

nm, as a result of the limited escape depth of electrons through a solid (33).

31

Typically, a low resolution scan from 0 to 1100 eV binding energy is performed
to determine the identity and concentration of each element present on the surface of a
sample, along with atomic ratios of elements. For wood products and polymers, carbon
and oxygen are often the primary elements of interest. The oxygen to carbon (O/C)
atomic ratio is calculated from the low resolution scan to evaluate oxidative changes
resulting from chemical reactions or weathering. To further analyze the chemical
bonding of the carbon atoms present in a sample, a high resolution scan of the C], region
from 280 to 300 eV is performed. F our carbon component peaks can be found from the
high resolution scan of wood products. Carbon component C1 arises from carbon atoms
bonded only to carbon and/or hydrogen atoms (C-C/C-H) and has a characteristic binding
energy of 285 eV. Carbon atoms bonded to a single oxygen atom, other than a carbonyl
oxygen (C-OH) constitute the C2 component, which has a binding energy of 286.5 eV.
Carbon component C3 arises from carbon atoms bonded to two non-carbonyl 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 non-carbonyl group (O-C=O).9’”'l7 The binding

energies of the C3 and C4 components of carbon are 288 and 289 eV, respectively (33).

XPS has been used to determine the chemical changes resulting from grafting
lignocellulosic fibers with maleated polyolefins (13-15, 17, 18) and to follow surface
changes to wood plastic composites after accelerated weathering (34, 35). It has also
been used to document surface chemistry changes in wood pulp after ozonation reactions

(36, 37).

32

Fourier Transform Infrared (F T IR) Spectroscopy

In FTIR spectroscopy, a beam of infrared radiation is used to excite the samples.
As a result, infrared radiation is absorbed by organic compounds and converted into
vibrational energy (38, 39). The mid IR region (4000 — 400 cm") is most commonly
used for analysis of chemical compounds. FTIR can be used to determine the presence of
functional groups in liquids or solids, using a variety of sampling methods. For solids, a
quantitative method of sample analysis is diffuse reflectance infrared Fourier transform
(DRIFT) spectroscopy (39, 40), in which the sample is finely ground and often mixed
with a non-absorbing substance such as potassium bromide (40). The method produces a
spectrum with bands corresponding to functional groups on the surface of the sample.
The spectra are presented in Kubelka-Munk units, which relate sample concentration to
the intensity of the bands in the spectra (40). The Kubelka—Munk equation is expressed

as:

_ 2
(1 R) :5 (1)

KR) : 2R 3

 

Where R is the absolute reflectance of the layer, k is the molar absorption coefficient and
s is the scattering coefficient (40).

Within the last two decades, FTIR has been used to study wood structure (41). It
has been applied to document the changes in the surface of wood after modification with
maleic anhydride (23, 26, 27, 29, 30), and to study the esterification reaction between

wood and maleated polyolefins (13, 15-17). FTIR has also been used extensively to

33

monitor changes in wood composites as a result of accelerated weathering (34, 35, 42-

44).

Nuclear Magnetic Resonance flVMR) Spectroscopy

NMR spectroscopy uses a magnetic field to energize the nuclei of selected atoms.
Samples with certain nuclei absorb electromagnetic radiation in the radio frequency
region (38). The absorption peaks for these nuclei are plotted versus peak intensity in the
NMR spectrum. While several atoms can be used for NMR spectroscopy, the two most
common forms, 1H and '3 C, are based on hydrogen and carbon nuclei, respectively (38).
Both of these methods provide information about bonding and chemical connectivity in
samples, though each has different advantages. For example, the integrated area under
each peak in the 1H spectra correlates well with the number of protons associated with
each peak (3 8). In ”C NMR, the peak areas do not typically correlate with the number of
carbon atoms present, so this technique is not quantitative unless run under special
conditions. However, each peak in the spectra usually represents a different form of
carbon, that is, a different bonding structure (3 8). For example, a carbon atom with two
attached hydrogen atoms will have a different chemical shift than carbon with an attached
hydroxyl group. Solid state 13C NMR is a useful technique for analysis of solid
materials. Solid state '3 C NMR has been used to study cellulose and other lignocellulosic
materials and their compounds (45-47). This technique has also been applied to
analyzing the carbon structure in wood (45, 46, 48). Solid state '3 C NMR was recently
used to determine grafting mechanisms in maleic anhydride grafted polypropylene

through reactive extrusion (49).

34

 

“ll:

\\0(

lllt‘ r

 

T itrimetric Analysis

Titrimetric analysis uses acid-base chemistry to determine the quantity of acidic
or basic groups in a substance. Techniques have been developed for the detection of acid
value (AV), saponification value (SV) and hydroxyl group content (HV) in alkyd resin
samples (50). These analyses have been applied to wood and other cellulosic samples for
the detection of changes in modified and grafted wood (1 3, 23, 26, 27, 29).

Acid value determines the amount of free carboxylic acid groups present in a
sample, while saponification value measures both acid and ester groups in a sample.
Hydroxyl value quantifies the free and accessible hydroxyl groups present in the sample

(50). These quantities were calculated using the following formulas:

AV = 56.1 1‘1. (2)
p
sv = 56.1(b—a)E (3)
p
HV = [(b—a)N (56.1) + AV (4)

where N is the normality of potassium hydroxide (KOH), p is the amount in grams of
wood particles, b is the volume of KOH needed to neutralize the blank (solvent only) and
a is the volume of KOH required to neutralize the sample. The factor 56.1 accounts for

the molecular weight of KOH (50).

35

AV and SV can provide information about esterification in samples , which is
desirable for wood modification studies (13, 23, 26, 27, 29). Hydroxyl group content is
also important in analyzing the way samples react (13). If the OH value is reduced after

modification, there is evidence that the reaction took place through these groups.

Physical and Mechanical Property Testing

Wood composite panels are classified according to their density as low, medium
or high density products (4, 6). To determine whether formaldehyde-free panels made in
. this study would conform to standard strength requirements for conventional
particleboard, density of the panels was assessed and mechanical property data was

compared to ANSI standard requirements for particleboard (51 ).

Density

Density of wood and wood-based samples is one of the most important
parameters in assessing sample quality (3, 4, 6). Since mechanical property requirements
vary with density, the density of experimental panels must be assessed to determine the
appropriate range of standard property values for product comparison (51). A simple
mass over volume calculation can be used to determine the average density. However,
more information about the density profile can be collected using an X-ray density
profiler. This instrument sends an x-ray beam through a sample as the beam travels
across the sample thickness. Results of this analysis show the variation in density across
the sample and can be used to find areas of high and low density in the sample, which

may account for mechanical property variations.

36

 

5F

de

In.

Sill

F lexural Properties

The flexural strength (modulus of rupture or MOR) and flexural stiffness
(modulus of elasticity or MOE) of composites are determined via static bending tests.
These tests are usually performed on dry samples, conditioned as per ASTM standards
before testing. The MOR and MOE can be calculated from the load-deflection curves

using the following equations:

MOR=————-—3'P""=”"L (5)
2-b-d2
3
MOE=—-3—L3— (6)

where Pmam is maximum load (N), P is the load at proportional limit (N), L is length of
span (mm), b is specimen width (mm), d is specimen thickness (mm) and y is center

deflection at proportional limit load (mm).

Internal Bond (IB) Strength

The internal bond (IB) strength measures the tensile strength perpendicular to the
surface of the sample and indicates how well the particles are bonded to each other in the
center of the panel. It is measured as the maximum load per unit area. In this test, the
stress is applied perpendicular to the plane of the adhesive and the energy required to

break the adhesive bond is measured. 18 strength is calculated according to Equation 7:

37

IB = —Pmax (7)

where Pmax is maximum load (N), b is the width of the specimen (mm) and d is the depth

of the specimen (mm).

38

 

J

14.

 

10.

ll.

12.

13.

14.

REFERENCES

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

Sjostrom, E., Wood Chemistry: fundamentals and applications, 2nd Ed.,
Academic Press: San Diego (1993).

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

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

Guss, L.M., "Engineered wood products: the future is bright," Forest Products
Journal, 45 (7/8): 17-24 (1995).

Maloney, T.M., Modern Particleboard and Dry-Process F iberboard
Manufacturing, Updated Ed., Miller Freeman: San Francisco (1993).

Sellers, T., Jr., "Wood adhesive innovations and applications in North America,"
Forest Products Journal, 51 (6): 12-22 (2001).

Maloney, T.M., "The family of wood composite materials," Forest Products
Journal, 46 (2): 19-26 (1996).

Pizzi, A., Wood Adhesives: Chemistry and Technology. Marcel Dekker: New
York (1983).

US-EPA, "National emission standards for hazardous air pollutants: plywood and
composite wood products," Federal Register, 69 (146): 45943-46046 (2004).

Clemons, C., "Wood-plastic composites in the United States: The interfacing of
two industries," Forest Prod. J., 52 (6): 10-18 (2002).

Principia Partners, Current and Emerging Applications for Natural and Wood
Fiber Composites, presented at 7th International Conference on Woodfibre-
Plastic Composites. 2003: Madison, WI.

Felix, J .M. and Gatenholm, P., "The nature of adhesion in composites of modified
cellulose fibers and polypropylene," Journal of Applied Polymer Science, 42 (3):
609-20 (1991).

Matuana, L.M., Balatinecz, J.J., Sodhi, R.N.S., and Park, C.B., "Surface
characterization of esterified cellulosic fibers by XPS and FTIR Spectroscopy,"
Wood Science and Technology, 35(3): 191-201 (2001).

39

 

‘1‘)

fib-

76.

 

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

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

Kazayawoko, M., Balatinecz, J.J., and Woodhams, R.T., "Diffuse reflectance
Fourier transform infrared spectra of wood fibers treated with maleated
polypropylenes," Journal of Applied Polymer Science, 66 (6): 1163-1173 (1997).

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

Kazayawoko, M., Balatinecz, J .J ., and Sodhi, R.N.S., "X-ray photoelectron
spectroscopy of maleated polypropylene-treated wood fibers in a high-intensity
thermokinetic mixer," Wood Science and Technology, 33 (5): 359-372 (1999).
Keener, T.J. and Brown, T., "Epolene maleated polyethylene coupling agents," 7’”
International Conference on Woodfiber-Plastic Composites, Madison, WI, May
19-20, 2003 : 23-27 (2004).

Selke, SE. and Wichman, 1., "Wood fiber/polyolefin composites," Composites,
Part A: Applied Science and Manufacturing, 35A (3): 321-326 (2004).

Balasuriya, P.W., Ye, L., Mai, Y.W., and Wu, J ., "Mechanical properties of wood
flake-polyethylene composites. II. Interface modification," Journal of Applied
Polymer Science, 83 (12): 2505-2521 (2002).

Matsuda, H., "Thermal plasticization of lignocellulosics for composites," in
Emerging Technologies for Materials and Chemicals fiom Biomass R.M. Rowell,
T.P. Schultz, and R. Narayan, Editors. American Chemical Society: Washington,
DC. p. 98-114 (1992).

Matsuda, H., "Preparation and utilization of esterified woods bearing carboxyl
groups," Wood Science and Technology, 21 (1): 75-88 (1987).

Matsuda, H., Ueda, M., and Mori, H., "Preparation and crosslinking of
oligoesterified woods based on maleic anhydride and allyl glycidyl ether," Wood
Science and Technology, 22 (1): 21-32 (1988).

Clemons, C., Young, RA, and Rowell, R.M., "Moisture sorption properties of
composite boards from esterified aspen fiber," Wood and Fiber Science, 24 (3):
353-63 (1992).

Timar, M.C., Mihai, M.D., Maher, K., and Irle, M., "Preparation of wood with
thermoplastic properties. Part 1. Classical synthesis," Holzforschung, 54 (1): 71-
76 (2000).

40

 

37

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

Timar, M.C., Maher, K., Irle, M., and Mihai, M.D., "Preparation of wood with
thermoplastic properties. Part 2. Simplified technologies," Holzforschung, 54 (1):
77-82 (2000).

Timar, M.C., Maher, K., Irle, M., and Mihai, M.D., "Thermal forming of
chemically modified wood to make high-performance plastic-like wood
composites," Holzforschung, 58 (5): 519-528 (2004).

Hon, D.N.S. and Xing, L.M., "Thermoplasticization of wood. Esterification," in
Viscoelasticity of Biomaterials, W.G. Glasser and H. Hatakeyama, Editors.
American Chemical Society: Washington DC. p. 118-132 (1992).

Marcovich, N.E., Reboredo, M.M., and Aranguren, M.I., "Sawdust modification.
Maleic anhydride chemical treatment," Holz als Roh- und Werkstojf, 54 (3): 189-
193 (1996).

Marcovich, N.E., Aranguren, M.I., and Reboredo, M.M., "Modified woodflour as
thermoset fillers Part I. Effect of the chemical modification and percentage of
filler on the mechanical properties," Polymer, 42 (2): 815-825 (2001).

Marcovich, N.E., Reboredo, M.M., and Aranguren, M.I., "Modified wood flour as
thermoset fillers. 11. Thermal degradation of wood flours and composites,"
Thermochimica Acta, 372 (1-2): 45-57 (2001).

Ratner, BB. and Castner, D.G., "Electron Spectroscopy for Chemical Analysis,"
in Surface Analysis : The Principal Techniques, J.C. Vickerrnan, Editor. John
Wiley & Sons: Chichester, England. p. 43-98 (1997).

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

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

Chtourou, H., Riedl, B., and Kokta, B.V., "Surface characterizations of modified
polyethylene pulp and wood pulps fibers using XPS and inverse gas

chromatography," Journal of Adhesion Science and Technology, 9 (5): 551-74
(1995)

Chtourou, H., Riedl, B., Kokta, B.V., Adnot, A., and Kaliaguine, 8., "Synthetic
pulp fiber ozonation: an ESCA and FTIR study," Journal of Applied Polymer
Science, 49 (2): 361-73 (1993).

Silverstein, RM. and Webser, F .X., Spectrometric Identification of Organic
Compounds, 6th Ed., John Wiley & Sons, Inc. : New York (1998).

41

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

Pemble, M.E., "Vibrational Spectroscopy from Surfaces," in Surface Analysis:
The Principal Techniques, J.C. Vickerman, Editor. John Wiley & Sons:
Chichester, England. p. 267-311 (1997).

Spectra-Tech, "The complete guide to FT—IR," 12.1-12.6 (1996).

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

Stark, N.M., Matuana, L.M., and Clemons, C.M., "Effect of processing method on

surface and weathering characteristics of wood-flour/HDPE composites," Journal
of Applied Polymer Science, 93 (3): 1021-1030 (2004).

Stark, NM. and Matuana, L.M., "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).

Matuana, L.M., Kamdem, D.P., and Zhang, J., "Photoaging and stabilization of

rigid PVC/wood fiber composites," Journal of Applied Polymer Science, 80 (11):
1943-1950 (2001).

Maunu, S.L., "NMR studies of wood and wood products," Progress in Nuclear
Magnetic Resonance Spectroscopy, 40 (2): 151 -174 (2002).

Hatfield, G.R., Sardashti, M., and Maciel, G.E., "Analysis of molecular
orientational order in solid samples by nuclear magnetic resonance: application to
lignin and cellulose in wood," Analytical Chemistry, 59 (13): 1659-64 (1987).

Jandura, P., Kokta, B.V., and Riedl, B., "Fibrous Long-Chain Organic Acid
Cellulose Esters and Their Characterization by Diffuse Reflectance FTIR
Spectroscopy, Solid-State CP/MAS 13C-NMR, and X-Ray Diffraction," Journal
of Applied Polymer Science, 78: 1354-1365 (2000).

Solum, M.S., Pugmire, R.J., Jagtoyen, M., and Derbyshire, F., "Evolution of
carbon structure in chemically activated wood," Carbon, 33 (9): 1247-1254
(1995)

Bettini, S.H.P. and Agnelli, J.A.M., "Grafting of maleic anhydride onto
polypropylene by reactive extrusion," Journal of Applied Polymer Science, 85
(13): 2706-2717 (2002).

Anonymous, "Recommended methods for the analysis of alkyd resins," Pure and
Applied Chemistry, 33 (2-3): 411-435 (1973).

ANSI, A208.I-I999. Particleboard. The Composite Panel Association:
Gaithersburg (1 999).

42

CHAPTER 3

COMPOSITE MATERIALS MANUFACTURED FROM WOOD PARTICLES

MODIFIED THROUGH A REACTIVE EXTRUSION PROCESS

This chapter is slightly modified from Polymer Composites, published in August 2005.
26 (4): 534-541. It is co-authored by K. Carlbom and L.M. Matuana.

43

ABSTRACT

Wood-based composites such as particleboard and medium density fiberboard are
currently made with formaldehyde-containing adhesives. Since the government is
continuously developing and implementing very stringent regulations to eliminate
formaldehyde emissions into the environment, alternative approaches must be developed
to replace these adhesives. This study examined the concept of using a reactive extrusion
process as a means of developing a new, formaldehyde-free binding system for wood
composite products. The surfaces of wood particles were modified by grafting maleated
polyolefins through a continuous reactive extrusion process. Chemical changes resulting
from this treatment were followed by studying the FTIR, '3 C NMR and XPS spectra.
The modified wood particles were compression-molded into panels, which were tested
for mechanical properties. FTIR, 13C NMR and XPS data revealed that chemical
reactions had taken place between the hydroxyl groups of wood particles and maleated
polyolefins. The mechanical property test results indicated that the composite panels

compared favorably with current standard requirements for conventional particleboard.

44

INTRODUCTION

Wood-based composites are commonly made using formaldehyde-based
adhesives, including urea-formaldehyde, melamine-formaldehyde, and phenol-
formaldehyde (1-3). In 1998, 1,780 kilotons of adhesive resin solids were used to
produce primary glued wood products (excluding the adhesive used to bond furniture and
other secondary wood products). Of this amount, nearly 92% were formaldehyde-based

adhesives (3).

Plants that produce wood composites using formaldehyde-based adhesives emit
harmful chemicals to the environment. These include phenol, formaldehyde, ketones,
and other compounds, which are known hazardous air pollutants (HAPs) (4). Common
composite products such as plywood, oriented strandboard and particleboard are used in
building construction and in furniture, which is a concern as these products also tend to
release formaldehyde over time (5). Wood composites made today emit far less
formaldehyde than those made 20 years ago, but the problem has not been eliminated.
Formaldehyde and other toxic compounds may be present in large amounts in both

indoor and outdoor air as a consequence of the use of these adhesives (6, 7).

As a result of public concern about the environment, the Environmental
Protection Agency (EPA) proposed new rules for facilities that manufacture plywood and
composite wood products in August of 2002 (6). If adopted, these rules would affect
both new and existing plants that produce at least 10 tons of any one HAP per year, or
any combination of 25 tons of HAPs per year. The California Air Resources Board has
gone even further; proposing a regulation that would eliminate urea-formaldehyde based

wood composites from being sold in California, regardless of where they were made (7).

45

 

or.
M
1hr
301'

up

5X11)

ll ()0

F00.

SPuL

 

These regulations will force industry to find new ways to bind composite products

without the use of formaldehyde-based adhesives.

In recent years, there have been several studies into environmentally friendly
wood adhesives (3). Some of the areas that have been investigated include urea-
formaldehyde adhesives with low forrnaldehyde-to-urea molar ratios, and the
development of tannin, lignin, soybean and cornstarch adhesives (3, 8), and phenol-
forrnaldehyde resins modified with lignin (9, 10). Additives that reduce formaldehyde
release during composite pressing and during board use have also been developed (8).
The regulation of formaldehyde emissions has also lead to some development of
fiberboard without synthetic adhesives (11, 12). The binderless boards and those made
with natural adhesives tend to have poorer properties than those made with synthetic

adhesives.

The approach of this work is to graft maleated polyolefins to wood particles in
order to bond the wood particles together without the use of additional adhesive. Prior
A work demonstrated the ability to graft maleated polyolefins to cellulosic materials
through a wet process (13). However, the wet process had the drawback of using organic
solvents, which had to be removed through drying. Therefore, the wet process is both

expensive and time consuming on an industrial scale.

The main objective of this study was to study the concept of using a reactive
extrusion process as a means of developing a new, formaldehyde-free binding system for
wood composite products. The effectiveness of the modification was followed by
Fourier transform infrared (FTIR) spectroscopy, nuclear magnetic resonance (NMR)

spectroscopy and X-ray photoelectron spectroscopy (XPS). FTIR is useful in

46

determining the presence of functional groups. NMR can determine bonding and
chemical connectivity, while XPS can reveal the elemental composition on the surface of
materials (13-15). Panels were pressed from the modified wood particles and mechanical
properties of the resulting panels were tested and compared with current standard

requirements for conventional particleboard (16).

47

EXPERIMENTAL
Materials

Maple wood particles of 425 micron (40-mesh) and 150 micron (100-mesh) size
were supplied by American Wood Fibers (Schofield, WI) and were used as particles. The
150 micron particles were used for the analytical work because the diffuse reflectance IR
technique required very small particles to minimize the effects of scattering and specular
reflectance in the samples. However, these small particles were difficult to feed into the
extruder. Since panel manufacturing required a large quantity of modified particles,
larger (425 micron) particles, which were easier to process, were used in panel
manufacturing and mechanical property testing. Hydrated zinc acetate, the catalyst, and
xylene (99.9%, ACS Grade), the solvent used for Soxhlet extraction, were obtained from
Baker Analytical Reagents (JT Baker Co., Phillipsburg, NJ). Maleated polyethylene (G-
2608 or MAPE) and maleated polypropylene (G-3003 or MAPP) supplied by Eastman
Chemical Co. (Kingsport, TN) were used as the binding agents. Table 3.1 lists the
characteristics of these binding agents. The wood particles were dried for 48 hours at
105°C to a final moisture content of less than one percent before processing. All other

chemicals were used as received.

48

Table 3.]. Characteristics of Maleated Polyolefins

 

 

Properties Maleated polyethylene Maleated polypropylene
(MAPE or G-2608) (MAPP or G-3003)
Maleic anhydride content 1.5% 1.5%
by weight
Melting point 122°C 156°C
Weight average molecular
weight (Mw) 51,700 g/mol 52,000 g/mol
Melt flow indexI 8 g/ 10 min 12.7 g/ 10 min

 

lMelt flow index measured at 190°C and 2.16 kg according to ASTM D1238.

49

Reactive Extrusion of Wood Particles

A 10-liter high intensity mixer (Papenmeier TGAHKZO) was used for dry
blending of the wood particles, binding agent, and catalyst. All components were
combined in the mixer and blended for 10 minutes at room temperature. Amounts of all
components used in the formulation are summarized in Table 3.2. The mixture was then
fed into a 32 mm conical counter rotating twin-screw extruder (C. W. Brabender
Instruments, Inc.) with a L/D ratio of 13:1, driven by a 7.5 hp Intelli-Torque Plasti-
Corder Torque Rheometer®. Based on preliminary work, the barrel temperatures for the
three zones inside the extruder were set at 160°C for maleated polyethylene and 165°C
for maleated polypropylene, and the rotational speed of the screws was held at 60 rpm.

No die was used to extrude these particles.

50

Table 3.2. Formulation Used for Surface Modification of Wood Particles

 

% Total in

 

Ingredients Composite Weight (g)
Wood particles 79% 790
Maleated polyolefins 20% 200
Hydrated zinc acetate 0
(catalyst) 1 /° 10

 

51

 

E1

5111
M;

U111

00
sec
hm

SEC

Su

1(15

(Ni

63C.

SUI):

and

11))

Extraction of Wood Particles

Modified and unmodified (unextruded) wood particles used for spectroscopic
studies were Soxhlet extracted with xylene following the approach described by Li and
Matuana (13). Particles were extracted for 24 hours after modification to remove any
unreacted binding agent, oven-dried at 105°C until constant weight was achieved, and
were then analyzed by FTIR. A second 24-hour Soxhlet extraction was then performed
to make sure the removal of unreacted binding agent was complete from the surface of
wood particles upon the first extraction. The infrared spectra of wood particles after this
second extraction were collected for comparison with those collected after the first 24-
hour extraction. All NMR and XPS analyses were performed on the particles after the

second extraction.

Surface Characterization of Wood Particles

Unmodified and modified wood particles were dried to a constant weight at
105°C and analyzed by infrared spectrophotometry, using a Nicolet Protége’ 460 FTIR
(N icolet Instrument Co., Madison, WI). Spectra were recorded in Kubelka-Munk units in
the range of 4000-400 cm", with a resolution of 4 cm'1 and a coaddition of 128 scans for
each spectrum. Pure powdered potassium bromide (KBr) was used as a reference
substance while no dilution of powdered-wood particles in KBr was required to obtain a
spectrum. Diffuse reflectance was used with the FTIR for transfer of infrared radiation
and data analysis was performed using WinFIRST software (Thermo-Nicolet, Madison,

WI).

52

 

obi.
0n 2
0pm
prol
amb
optii
ICCyi

wig

and
surf;
on

C1111

The

Pan

PICS:
large
Press

It’mpc

Solid-state carbon 13 nuclear magnetic resonance (”C NMR) spectra were
obtained using a 1H-13 C Cross-Polarization Magic Angle Spinning (CP-MAS) experiment
on a Varian-Chemagnetics (Varian Inc., Palo Alto, CA) 400 MHz NMR Spectrometer
operating at 100.529 MHz. The spectrometer was equipped with a double resonance
probe. Samples were spun in a 6 mm rotor at 4 kHz, at the magic angle (54.7°) and at
ambient temperature. Pulse widths, contact time and pulse amplitudes for the CP were
optimized on an adamantane standard. The contact time in all cases was 1 ms. The
recycle delay (pulse delay) was 2.0 s. Spectra were processed using exponential

weighting (100 Hz).

X-ray photoelectron spectroscopy (XPS) was used to determine the concentration
and types of carbon atoms, as well as the oxygen-to-carbon atomic ratios present on the
surface of the wood particles before and after modification. XPS analysis was carried out
on a Physical Electronics Phi 5400 ESCA System, (Physical Electronics USA,
Chanhassen, MN) using a non-monochromatic Mg source and a takeoff angle of 45°.

The procedure for XPS data collection and analysis was detailed in other articles (13, 14).

Panel Manufacturing and Mechanical Property Testing

Panels were prepared from modified 425 micron wood particles using a hydraulic
press from Eric Mill Co. (Erie, PA). Panel dimensions were 380 by 380 by 6 mm, with a
target density of 720 kg/m3. Panels were pressed at 193°C for 7 minutes using 8 MPa of
pressure. After pressing, panels were removed from the hot press and cooled at room

temperature under compression for 15 minutes.

53

test
test:
D1 I
of n

with

Stat

dete
b011-

Ex;

Three-point flexural tests and internal bond (IB) strength tests were performed on
an Instron 4206 testing machine (using Series IX software). For three-point flexural
tests, the crosshead speed was 3.05 mm/min, while the crosshead speed for 18 strength
tests was 8.13 mm/min in accordance with the procedure outlined in ASTM standard
D1037-99 (17). At least six samples were tested to obtain an average value for modulus
of rupture (MOR), modulus of elasticity (MOE), and 1B strength, which were compared

with the values listed in the standard ANSI A208 . l -1 999 Particleboard (16).

Statistical Analysis

A two-sample t-test was carried out with an a significance value of 0.05 to
determine the effect of binding agent type (MAPE vs. MAPP) on the flexural and internal
bond properties of the composites. Statistical analysis was performed using Design

Expert software (Version 6) from Stat-Ease. Inc. Minneapolis, MN.

54

50

met

thr0

char

will

RESULTS AND DISCUSSION
Surface Characterization of Wood Particles

Modification of the wood particles was expected to take place through the
mechanisms proposed in Figure 3.1, where a monoester or a diester could be formed
through reaction of the maleated polyolefin with hydroxyl groups on the wood surface
(13-15). Various surface characterization techniques were then used to document
changes in the chemistry of the unmodified wood particles and wood particles modified

with maleated polyolefins through a reactive extrusion process.

55

 

 

Figt

 

O
fill—OH 0%?‘2
I “ (R).
/' OH CH
[— o>c \ /

i‘R'
(R)n
Wood
particle Maleated
polyolefin
*CH2_ CH2_)'_
Where R = or
CH3
Figure 3.1.

_ CH3

l 8
/|—O—C—CH2

3‘01'1 />3H /(Rln
| HO’fi \C-R' Monoester
o l
(Rln
| 8
/j—O—C—CH2
/ I
/l—O—E—CH (Rln
/' 0 \C/—R' Di-ester

Modified wood (Rln
particle

Modification scheme for esterification reaction between wood particles

and maleated polyolefins: (a) mono-ester and (b) di-ester formation.

56

1118

sh0

sh0‘

strel

\‘ibr

£11101

C0111

 

The FTIR spectra of unmodified and modified wood particles, along with the
maleated polyethylene are shown in Figure 3.2. Similarly, the FTIR spectra of
unmodified and modified wood particles, along with the maleated polypropylene are
shown in Figure 3.3. The unmodified wood particles (spectrum A in Figures 3.2 and 3.3)
showed an absorbance band at 3400 cm", which was attributed to hydroxyl group
stretching vibrations, and another at 2900 cm", which was associated with C-H stretching
vibrations. A band near 1740 cm'1 was associated with C=O stretching vibrations, and
another at 1122 cm“1 was likely due to C-0 stretching vibrations and C-C stretching from

components of cellulose (13, 15).

57

rang-mx—oacx

inn—3C

Figure 3.2.

 

 

 

 

 

N
3% N
.9:
D NN
8‘53
N N
E?
E
4000 3500 3000 2500 2000 1500 1000 500

Wavenumbers (cmA-1)

FTIR spectra of unmodified wood particles (A), maleated polyethylene-
MAPE (B), MAPE-modified wood particles without extraction (C), with
24 hour-Soxhlet extraction (D), and after a second 24 hour-Soxhlet
extraction (E) in the region 4000-400 cm".

58

 

rang-mx—oacx

 

mnmzc

 

 

 

4000 3500 3000 2500 2000 1500 1000 500
Wavenumbers (cmA-1)

Figure 3.3. F TIR spectra of unmodified wood particles (A), maleated polypropylene-
MAPP (B), MAPP-modified wood particles without extraction (C), with
24 hour-Soxhlet extraction (D), and after a second 24 hour-Soxhlet
extraction (E) in the region 4000-400 cm".

59

Fig

Silt
1168

171

and

and

m

be
ex

P3.

EhSG

The spectra of maleated polyethylene and maleated polypropylene (spectrum B in
Figures 3.2 and 3.3, respectively) showed four distinct absorption bands. The bands
between 2960 cm'1 and 2840 cm'1 were due to symmetrical and asymmetrical C-H
stretching vibrations of CH2 and CH3 in the polyolefin chain (13, 15). The small band
near 1786 cm'l was attributed to anhydride C=O stretching. Bands found between 1720-
1710 cm'I arose from C=O stretching vibrations in the maleated polyolefins and the
absorbance bands from 1463 to 1300 cm'1 were from C-H deformation vibrations of CH2
and CH3. The bands below 1250 cm'1 were associated with rocking vibrations of CH2

and CH3, or C-C stretching vibrations from the polyolefin chain (13, 15).

The infrared spectra of wood modified with maleated polyolefins are also
illustrated in Figures 3.2 and 3.3 (spectra C-E). For the interpretation of infrared spectra,
it is important that all the unreacted maleated polyolefins are removed in the 24 hour
Soxhlet extraction with solvent. It is seen from these figures that there was very little
difference in the spectra of the modified wood particles after the first (spectrum D) and
second extractions (spectrum E). This finding provides important evidence that the
maleated polyolefins were chemically bonded to the wood particles. If the maleated
polyolefins had not grafted to the particles, a decrease in peak intensity or a loss of the
band with distinct peaks between 2960 and 2840 cm'1 would have been expected after the
extraction, which would remove any polyolefins not chemically bonded to the wood

particles (13).

The modified wood particles (spectra C-E in Figures 3.2 and 3.3) also showed
absorption bands at 3400 cm", but the intensity of this peak decreased compared to the

unmodified wood particles, indicating that there were less OH groups on the surface of

60

[1101‘
11111
6511
163

5111'

rep
the
rib:
incr
rear
in c
imci

1680

0:110

llj’dri

surfai
SPCCII
lSpeQi
Who

did 11

 

modified samples. This was expected based on the reaction scheme shown in Figure 1,
where the maleated polyolefin reacted with the OH groups of wood particles forming an
ester link (13). The reaction may form either a single ester link or a diester where both
reactive groups on the maleated polyolefin bond with hydroxyl groups on the wood
surface. A distinct change was found near 2900 cm'l, where a large band between 2960
and 2840 cm], having distinct peaks similar in appearance to the maleated polyolefin,
replaced the single peak in the unmodified wood particle spectrum. This feature is
characteristic of the maleated polyolefins (spectrum B), and is due to C-H stretching
vibrations (13, 15). Another indication of grafting of the maleated polyolefin was an
increased intensity in the band at around 1740 cm", possibly due to esterification
reaction. Although not clearly seen in Figures 3.2 and 3.3 due to stacking of five spectra
in one figure, the individual spectra of modified wood particles showed an increased
intensity in the peaks at 1462 and 1376 cm'1 which is also indicative of the grafting
reaction, suggesting more C-H character in the modified samples (13, 15). It should also
be mentioned that the intensity of the band at 1122 cm'I decreased, likely due to less
cellulose being detected on the surface because of the grafting of the long aliphatic

hydrocarbon chain of maleated polyolefin (13-15).

The evidence supporting the chemical bonding of maleated polyolefins to the
surface of wood particles obtained by infrared was further supplemented by NMR
spectra, as illustrated in Figure 3.4. The solid state '3 C NMR maleated polyethylene
(spectrum A) showed only one distinct peak at 30 ppm, which was due to CH2 chain
carbons of the maleated polyethylene (18). Oxidized carbons from the maleated group

did not appear, likely due to their low concentration in the maleated polyethylene

61

C011
“’01

1‘1)n

3101

(1‘1
mm
was
wet

spec

 

compounds. By contrast, spectra of both unmodified wood (spectrum B) and modified
wood particles (spectrum C) showed several characteristic peaks. The carbon peak at 152
ppm is attributed to lignin compounds in the wood. Bands from different types of carbon
atoms in cellulose appeared at 106 and 73 ppm. Wood particles also showed a small
peak near 55 ppm attributed to methoxy groups found in wood lignin and hemicelluloses
(19). As expected, a large increase in the peak intensity at 30 ppm was observed in the
modified wood particles (spectrum C), clearly indicating that the maleated polyethylene
was chemically bonded to the wood particles. Any unbound maleated polyethylene
would have been removed by the Soxhlet extraction process, and would not appear in the

spectrum.

62

 

 

 

Figu

 

3O

 

 

 

 

 

 

55
152
B _ __ 73
106
c —t
300 250 200 150 100 50 0 PPM

Figure 3.4. Solid-state 13C NMR spectra of (A) MAPE, (B) unmodified wood
particles, and (C) wood particles modified with MAPE.

63

FTIR
signi
male
signi
\l'lllC
surfa
511113
0.47
pols

was ‘

The XPS data summarized in Table 3.3 also confirmed the findings of the
FTIR analysis. As expected from compounds rich in carbon-containing groups, a
significant increase in C1 was observed after surface modification of wood particles with
maleated polyolefins. In addition, the content of oxidized carbon atoms (C2-C4)
significantly decreased in the modified wood particles, along with the O/C atomic ratio,
which was also expected due to the large increase in aliphatic carbon atoms on the
surface of the modified wood particles (13, 14). The O/C atomic ratios showed that the
surface of the wood particles changed dramatically with modification, decreasing from
0.47 to 0.03 with maleated polyethylene and from 0.47 to 0.06 with maleated
polypropylene. Changes this large likely indicate that the reactive extrusion procedure

was very successful in grafting maleated polyolefins to the wood particles.

64

 

Am: AOHUAVV 938w 38083-80: m was 388.88 a 2 B8: 08 £023

V803 conned 80¢ $88 cab 8088 v0 88 AOHU .O-U-Ov 89m nowxxo 3:088 Bwfim a 8 do .0808 cowxxo 78383
-80: 95 8 causes m88m 82:8 80¢ momma MU .AEOOV samba 38588 a 82: 550 .803 cowtmxo Bwfim a 8 Becca .58on
8588 80¢ m0 AIOOOV m88m come??— cohoce 8580 9 Eco wagon m89e 85:3 89a memes _ D 88388 :onBUN

823868 628m s cm @883 a Home 38890 39: @803 com: 8383 too? uofiuoE van 3588:?

 

 

 

. . . . . . . . 2:22
8 o 2 ea 3 m o o g N 2 e so as as, BEBE 83:8 895
8d See mam ed 33 EN whee 6.23: 828283; Base:
. . . . . . . $22
8 o 8 3 8 N o o mm o S e we we as) eeceee 3.8.8 895
:.o 0.8 2: o.o am I Em GEE 822338 Base:
. . . . . . . 8825.23
5. o 2 we a x mm o 8 m cm a me a 83:8 noes eeeeeea:
8:2 o o 3 8 8 6
383a

Ae\ev 8352—83 Viv—3:2

US .3585 ”as; 9.8.. .6 he were:

 

max 43 8880on $35.82* woo? mo 80989800 888m 380805 use 33m 20 cots—cmomhwi .m.m 2an

65

)lecl

bind"
as H
belu
11001
expe
as 51
woul
Ipor
path

311) at

 

 

 

Mechanical Properties

Surface analysis techniques provided evidence that the maleated groups of the
binding agent had chemically bonded with hydroxyl groups on the wood particle surface,
as illustrated in Figure 3.1. This process resulted in the formation of an ester linkage
between the wood surface and the maleated group of the binding agent, which produced a
wood particle with a pendant polyolefin chain. Bonding between wood particles was
expected to occur through diffusion and entanglement of the pendant polyolefin chains,
as shown in Figure 3.5. During compression molding, the pendant polyolefin chains
would melt and flow under the influence of heat and pressure, causing them to entangle.
Upon cooling, the entangled polyolefin chains would be locked together, forming a direct
particle-to-particle bond between the modified wood particles, without the addition of

any adhesive.

66

Figiu

 

 

 

0 I

 

 

 

 

I (II ll
/j—O—C—CH2 Hzc—C—O—‘I\
él 0H CH /(R)n (Rln\ CH H0 I?

/
' HO’fi \C—R' + R—C/ I‘OH '
O l l o
(R)n (Rln
Modified wood particle 2 Modified wood particle
i
Panel
Figure 3.5. Panel manufacture scheme for modified wood particles where R is an

67

ethylene or propylene repeat unit and R’ is hydrogen or a methyl group.

5116

C011

6%

C311

C011

0111C

 

 

To determine whether panels made by this process would conform to standard
strength requirements for conventional particleboard, mechanical property data was
compared to ANSI standard requirements for particleboard. Table 3.4 lists the MOR,
MOE, and 18 strength requirements for particleboard of medium density, ranging from
640-800 kg/m3. There are four grades of particleboard of medium density, all of which
can be made with either interior or exterior adhesives. Grades M-1 and M-S are
commercial grade boards, while M-2 and M-3 are intended for industrial use. Panels for
outdoor use must also be labeled exterior, according to the ANSI standard for
particleboard (16). The composite panels manufactured in this study were within this
range, with an average density value of 778 i 12 kg/m3 for panels made with maleated

polyethylene and 775 i 8 kg/m3 for panels made with maleated polypropylene.

68

Table

SIC
111

11

l
Fion‘

3

'Ihei
denou
Chhcr

 

 

Table 3.4. Requirements for Various Grades of Particleboard of Medium Density
(640-800 kg/m3)

 

Gradesl Experimental Values2

 

Properties
M-l M-S M-2 M-3 Wood/MAPE Wood/MAPP

 

MOR (N/mmz) 11.0 12.5 14.5 16.5 20.7 2: 3.4 A 22.9 i 4.3 A
MOE (N/mmz) 1725 1900 2250 2750 1296 i 195 A 2870 i 320 B

[B (N/mmz) 0.40 0.40 0.45 0.55 2.06 i 0.7 A 1.50 2: 0.3 A

 

lFrom Standard ANSI A208.1-1999 Particleboard (16).

2The capitalized letters represent the ANOVA results. The means with the same letter
denote that the difference between these two treatments is not statistically significant.
Otherwise, the difference is statistically significant at values of or = 0.05.

69

of pa
been
signi
1101.“
513110
bindi
T1101
iOb‘i
101.11
Mi
10111
[381111
form;
6010

ANS

 

 

Regardless of the maleated polyolefin type, the MOR and IB strength results
of panels manufactured in this study indicated that the standard requirements have
been met and surpassed for all grades of particleboard of medium density. No
significant difference in both MOR and 18 was observed between maleated
polyethylene and maleated polypropylene. However, the MOE data are below the
standard requirements for stiffness when maleated polyethylene was used as a
binding agent for wood particles. By contrast, the experimental panels exceeded the
MOE requirements for all grades of particleboard of medium density when maleated
polypropylene was used as binding agent. Panels bonded with maleated
polypropylene were statistically superior in MOE compared to the maleated
polyethylene counterparts. This difference is due to the higher stiffness of
polypropylene when compared to polyethylene. These results are significant because
particleboard is currently manufactured with formaldehyde-based adhesives. The
formaldehyde-free wood composites manufactured in this study are more
environmentally friendly and often outperform the requirements listed in the standard

ANSI A208.].

70

p0ly0?
based
wood
could
withm
emim
wood

perior

requiic

 

 

CONCLUSIONS

This study examined the possibility of modifying wood particles with maleated
polyolefins in a reactive extrusion procedure in order to make formaldehyde-free wood—
based composite panels. FTIR, 13 C NMR and XPS results verified the reaction between
wood particles and maleated polyolefins. This proved that the maleated polyolefins
could be successfully grafted to wood particles using a reactive extrusion process,
without the use of any solvents. The study also showed that a new type of
environmentally friendly wood composite product could be formed from the modified
wood particles. This composite contained no formaldehyde-based adhesive, but still
performed very well in mechanical tests, in some cases exceeding the standard

requirements for particleboard of medium density.

71

Ix.)

SJ)

 

 

 

10.

11.

12.

REFERENCES

Maloney, T.M., "The family of wood composite materials," Forest Products
Journal, 46 (2): 19-26 (1996).

Guss, L.M., "Engineered wood products: the future is bright," Forest Products
Journal, 45 (7/8): 17-24 (1995).

Sellers, T., Jr., "Growing markets for engineered products spurs research," Wood
Technology, May/June: 40-43 (2000).

Barry, A., Comeau, D., and Lovell, R., "Press volatile organic compound
emissions as a function of wood particleboard processing parameters," Forest
Products Journal, 50 (10): 35-42 (2000).

Maloney, T.M., Modern Particleboard and Dry-Process F iberboard
Manufacturing, Updated Ed., Miller Freeman: San Francisco (1993).

Anonymous, “Summary of working draft of proposed rule for plywood and
composite wood products,” National Emission Standards for Hazardous Air
Pollutants (N ESHAP), Rule Development Project Lead: Greg Nizich

( nizichgrgg@epa. gov), US. EPA, Technology Transfer Network-Air Toxics
Website, August 2002, http://wvwv.epagov/ttn/atw/plypart/plypart.html

 

Anonymous, “Fact Sheet: Composite Wood Products,” California Air Resources
Board Website, March 2003, htth/wwwarbca. gov

Pizzi, A., Wood Adhesives .' Chemistry and Technology. Marcel Dekker: New
York (1983).

Matuana, L.M., Riedl, B., and Barry, A.O., "Kinetic characterization by DTA of
lignosulfate-based phenol-formaldehyde resins," European Polymer Journal, 29
(4): 483-90 (1993).

Kazayawoko, J .S.M., Riedl, 13., Poliquin, J ., Barry, A.O., and Matuana, L.M., "A
lignin-phenol-formaldehyde binder for particleboard. Part 1. Thermal
characteristics," Holzforschung, 46 (3): 257-62 (1992).

Velasquez, J .A., Ferrando, F., Farriol, X., and Salvado, J ., "Binderless fiberboard
from steam exploded Miscanthus Sinensis," Wood Science and Technology, 37

(3-4): 269-278 (2003).
Widsten, P., Qvintus-Leino, P., Tuominen, S., and Laine, J .E., "Manufacture of

fiberboard from wood fibers activated with F enton's reagent (H202/FeSO4),"
Holzforschung, 57 (4): 447-452 (2003).

72

13.

14.

15.

16.

17.

18.

19.

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

Matuana, L.M., Balatinecz, J.J., Sodhi, R.N.S., and Park, C.B., "Surface
characterization of esterified cellulosic fibers by XPS and FTIR Spectroscopy,"
Wood Science and Technology, 35 (3): 191-201 (2001).

Kazayawoko, M., Balatinecz, J .J ., and Woodhams, R.T., "Diffuse reflectance

Fourier transform infrared spectra of wood fibers treated with maleated
polypropylenes," Journal of Applied Polymer Science, 66 (6): 1163-1173 (1997).

ANSI, A208.1-1999, Particleboard. The Composite Panel Association:
Gaithersburg (1999).

ASTM, D 103 7-99, Standard Methods of Evaluating the Properties of Wood-
Based Fiber and Particle Panel Materials. ASTM: West Conshohocken (1999).

Heinen, W., Rosenmoller, C.H., Wenzel, C.B., de Groot, H.J.M., Lugtenburg, J .,
and van Duin, M., "13C NMR study of the grafting of maleic anhydride onto
polyethylene, polypropene and ethene-propene copolymers," Macromolecules,

29: 1151-1157 (1996).
Solum, M.S., Pugmire, R.J., Jagtoyen, M., and Derbyshire, F., "Evolution of

carbon structure in chemically activated wood," Carbon, 33 (9): 1247-1254
(1995)

73

FI‘

 

 

This

auth

 

CHAPTER 4

F UNCT IONALIZATION 0F WOOD PARTICLES THROUGH A REACTIVE

EXTRUSION PROCESS

This chapter is in press for the Journal of Applied Polymer Science (2006). It is co-
authored by K. Carlbom and L.M. Matuana.

74

P01)"
MA]
durii
were
parti
dil‘fe
XPS
ester
0f 1r.
ll}'dri
poly
deter
Signi
cond
no d
extru
male

studi

 

 

ABSTRACT

Wood particles were modified in a reactive extrusion process with maleated
polyethylene (MAPE) and maleated polypropylene (MAPP) compounds. Contents of
MAPE were varied to study the effect of material composition on grafting efficiency
during reactive extrusion, while extruder barrel temperatures and rotational screw speeds
were varied to evaluate the effects of processing conditions on the modification of wood
particles. Polymer molecular weight effects were investigated using MAPP with
different molecular weights. Efficiency of the modification was assessed using FTIR and
XPS surface analysis techniques, along with a titrimetric analysis to verify the
esterification reaction between the wood particles and maleated polyolefins. The grafting
of maleated polyolefins onto the surface of the wood particles through a reaction of
hydroxyl groups on the wood surface with the maleated groups of the maleated
polyolefins was confirmed, while the level of grafting of MAPE onto wood particles was
determined to be a function of the MAPE concentration. However, there was no
significant difference found in grafting efficiency at different extrusion processing
conditions; rather all of the conditions resulted in adequate grafting. Similarly, there was
no difference in grafting efficiency with the molecular weight of MAPP. Reactive
extrusion was found to be a suitable technique for the modification of wood particles with
maleated polyolefins for all of the material compositions and processing conditions

studied.

75

fhnn
pohi
pn0r
modi
smne
ihnn.

noth

High
hydn
thegi
SeVei
(0r0
fiber
05s
50h
fille.
lime
Duh

00m

INTRODUCTION

In response to the need for new adhesives for wood composite products, a
formaldehyde-free binding system for wood particles which used only maleated
polyolefins to create direct wood particle to wood particle bonding was developed in our
prior work (1, 2). Wood composite panels were manufactured from maleated polyolefin-
modified wood particles, without using any additional adhesive. These panels met and in
some cases even exceeded standard requirements for particleboard made with
formaldehyde-based adhesive. However, the mechanism for the adhesive bonding was

not firlly developed.

Maleated polyolefins react with wood particles through an esterification reaction
(Figure 4.1). This reaction occurs between the maleated groups of the polyolefins and
hydroxyl groups on the wood surface, forming a monoester or a diester. This results in
the grafting of a pendant polyolefin chain to the wood surface through the ester linkage.
Several authors have studied this reaction with maleated compounds and wood particles
(or other cellulosic materials) (3-14). However, most of this previous work focused on
fibers modified through a solvent-based process where the maleated compounds were
dissolved in heated xylene, toluene, N,N-dimethylformamide (DMF) or other organic
solvents and cellulosic particles were added (3-10, 12-14). Cellulosic fibers have to be
filtered and dried after modification, which makes this process both cost-ineffective and
time consuming, and the often harmful organic solvents must be properly disposed of. In
order to produce modified wood particles that could be used to make alternative

composite panels, it is necessary to change from a solvent-based process to a dry process.

76

o

I

/I_OH %CH2
A! OH + (3)/0'11 (R)"
f— 0 \ /

C-R'
(Rln
Wood
particle Maleated
polyolefin
—(CH2— CH2-)—
Where R = or
——(CH2— <le -)—
CH3
Figure 4.1.

_ CH3

OH CH (Rln
/ /
l C - R'

(Rln

Modified wood (Rln
particle

Monoester

Diester

Modification scheme for esterification reaction between wood particles

and maleated polyolefins: (a) monoester and (b) diester formation.

77

then

fibel

not

for

pro

for:

of

im

fox

ihl

6X

CS

C0

EX

mm

and

 

A solvent-free process that has been reported in the literature utilized a
thermokinetic high-intensity mixer to graft maleated polyprOpylene (MAPP) onto wood
fibers (11, 12). However, the esterification reaction of wood particles and MAPP could
not be confirmed through surface characterization techniques. An effective dry process
for grafting maleated polyolefins onto the surface of wood particles is still needed to
produce a large volume of modified wood particles required to manufacture

formaldehyde-free wood composite panels.

Reactive extrusion is a technique that can be used for the chemical modification
of compounds, usually polymers that can melt during processing. Because no solvent is
involved during processing, reactive extrusion is an environmentally friendly technique
for surface modification. It also allows for control of reaction times and temperatures
through variation of rotational screw speeds and extruder barrel temperatures. Reactive
extrusion offers the advantage of intensive mixing of components in a short time,
especially in a twin-screw extruder, which ensures a more thorough reaction of
components (15). This process also has the advantage of low retention times in the
extruder, so a series of different reaction conditions can be tested in a relatively short
period of time. Reactive extrusion has only rarely been used for wood fiber surface
modification (1, 2) probably because the wood component does not melt and flow in the
extruder. However, the ease of processing and control of reaction conditions makes this

technique attractive for wood fiber surface modification.

In this study, reactive extrusion was employed to graft maleated polyethylene
(MAPE) and MAPP compounds to wood particles, with an ultimate goal of determining

optimum surface modification conditions. Particular emphasis was placed on examining

78

 

the E
comp
exUtl
photc
finin‘
panic
apprt
scnsi

aboul

 

the efficiency of grafting of the surface modification as functions of material
compositions (MAPE contents and molecular weights of MAPP compounds) and
extrusion processing conditions (barrel temperatures and rotational screw speeds). X-ray
photoelectron spectroscopy (XPS), Fourier transform infrared (FTIR) spectroscopy and
titrimetric analysis were used to monitor chemical changes on the surface of the wood
particles before and after reactive extrusion with maleated polyolefins. XPS has an
approximate sampling depth of 0.1-8 nm (16), which makes it even more surface
sensitive than FTIR. When used together, these techniques provide detailed information

about the surface chemistry and bonding mechanism in modified wood fibers.

79

 

Slater

supplil
and X)
from

polyet
3316.
were
in Ta‘
NI).
butar
Bake
magi
metl‘
cont

lfiCC'

 

 

EXPERIMENTAL

Materials

Maple wood particles of 150 micron (lOO-mesh) size used in this study were
supplied by American Wood Fibers (Schofield, WI). Hydrated zinc acetate, the catalyst,
and xylene (99.9%, ACS Grade), the solvent used for Soxhlet extraction, were obtained
from Baker Analytical Reagents (JT Baker Co., Phillipsburg, NJ). Maleated
polyethylene (MAPE or G-2608) and several maleated polypropylenes (MAPP or G-
3216, G-3015, G—3003, and E43) supplied by Eastman Chemical Co. (Kingsport, TN)
were used as the binding agents. Characteristics of these maleated polyolefins are listed
in Table 4.1. Hydrochloric acid (concentrated) obtained from EM Science (Cherry Hill,
NJ), sodium hydroxide, methanol (absolute), pyridine (99.9%), toluene (100%), 1-
butanol (99.6%), ethanol (190 proof), and acetic anhydride (100%), all obtained from
Baker Analytical Reagents (JT Baker Co., Phillipsburg, NJ), were used as titration
reagents. Titration reagents were used as directed in Pure and Applied Chemistry
methods (17). The wood particles were dried for 48 hours at 105°C to a final moisture
content of less than one percent before processing. All other chemicals were used as

received.

80

Table‘

‘Wkfigr
‘ llehir
Appro
arerag
MW (

‘5

lleh l

    
  
   
  
  

\isco

——_—.

1Melt
2(3.3,:

 

 

Table 4.1. Characteristics of Maleated Polyolefins

 

 

 

 

. MAPE MAPPZ
Properties

G-2608 E-43 G-3216 G-3015 G-3003
Weight % maleic anhydride 1.5 8 2.5 2.5 1.5
Melting point (°C) 122 155 142 155 156
Approximate weight
average molecular weight 51,700 1 1,200 39,000 47,000 52,000
Mw (g/mol)
Melt flow index at 190°C1 8 ---- ---- ---- 12.7
Viscosity at 190°C, cP ---- 300 20,000 25,000 60,000

 

lMelt flow index measured at 190°C and 2.16 kg according to ASTM standard D1238.
2G-3216 is a maleated PE/PP copolymer.

81

Surfi
Extn

dn>l
confl
conu

inoz

uas<
hadti
speec
thee‘
toth

each
zone

{roe

 

 

Surface Modification of Wood Particles with Maleated Polyolefins in Reactive
Extrusion

A 10-1iter high intensity TGAHK20 mixer (Papenmeier, Germany) was used for
dry-blending of the wood particles, binding agent, and catalyst. All ingredients were
combined in the mixer and blended for 10 minutes at room temperature. Amounts of all
components used in the formulation are listed in Table 4.2. These mixtures were then fed
into a 32 mm conical counter-rotating twin-screw extruder (C. W. Brabender Instruments,
Inc., South Hackensack, NJ) with a length to diameter (L/D) ratio of 13:1. The extruder

®. The extruder

was driven by a 5.6 kW Intelli-Torque Plasti-Corder Torque Rheometer
had three temperature zones on the barrel (Figure 4.2) and an adjustable rotational screw
speed. Various processing temperatures and rotational screw speeds were used to study
the effects of processing conditions on the effectiveness of grafting maleated polyolefins
to the surface of the wood particles. In all cases, a uniform barrel temperature was set for
each test. For example, when the desired processing temperature was 130°C, all three

zones were set at that temperature. This was done similarly for the other studied

processing temperatures of 140, 160 and 180°C.

82

Table

Bind

 

 

Table 4.2. Formulations Used for Modified Wood Particles

 

Ingredients in the formulations

 

Binder content

 

 

(%) Maple(particles MAPE or MAPP Hytil’caettidezinc
g) (g) (g)
5 94 5 1
10 89 10 1
15 84 15 l
20 79 20 1
Hopper
Motor Gear 1 Zone 1 Zone 2 Zone 3

 

Figure 4.2. Diagram of the extruder showing the three heating zones.

83

Inoc
ban
Hnn
[em]

Vafit

Stud:
COIIC

Spec

Exn

rent

Wei

 

 

The effect of material composition on the grafting of MAPE onto wood particles
in a reactive extrusion process was investigated by holding the extruder barrel
temperature and rotational screw speed constant at 160°C and 60 rpm, respectively.
Whereas, the MAPE contents were varied from 5 to 20% of the total batch weight (Table

4.2).

To determine the effects of processing conditions on grafting efficiency of
modified wood particles, the MAPE concentration was fixed at 20% and the extruder
barrel temperatures and rotational screw speeds were varied. Barrel temperatures were
limited to the range above the melting point of MAPE and below the degradation
temperature of wood (130, 140, 160 and 180°C). The rotational speed of the screws was

varied from 20 to 80 rpm in 20-rpm increments at each barrel temperature profile.

The influence of molecular weight on the efficiency of the grafting reaction was
studied using four different MAPP compounds (Table 4.1). Extrusion processing
conditions were held constant with the barrel temperature at 160°C and the rotational

speed of the screws at 60 rpm.

Extraction of Wood Particles

Unmodified wood particles were Soxhlet extracted with acetone for 24 hours to
remove impurities, air dried for 60 hours, and then oven dried at 105°C to a constant
weight before using for further analysis. Modified wood particles were also Soxhlet
extracted with xylene for 24 hours to remove any unreacted maleated polyolefins which

had not been linked to the wood particles, followed by air drying for 60 hours, and oven

84

dryii

by i

first

dne

Sur

usin
dete
mod
cm'
spec
Pure
of t
perf

mm

“01

 

 

drying to a constant weight at 105°C. These modified wood particles were then analyzed
by FTIR. A second 24-hour Soxhlet extraction was performed to ensure the removal of
unreacted maleated polyolefin was complete from the surface of wood particles upon the
first extraction. Following the second extraction, modified wood particles were again air
dried for 60 hours and then dried to a constant weight at 105°C, after which FTIR, XPS

and titrimetric analyses were performed.

Surface Characterization of Wood Particles

FTIR spectra of both unmodified and modified wood particles were obtained
using a Nicolet Prote’gé 460 FTIR spectrometer (N icolet Instrument Co., Madison, WI) to
determine the functional groups present at the surface of the samples before and after
modification. Spectra were recorded in Kubelka-Munk units in the range of 4000-400
cm", with a resolution of 4 cm'1 and a coaddition of 128 scans for each spectrum. All
spectra were collected using diffuse reflectance with FTIR for transfer of IR radiation.
Pure powdered potassium bromide (KBr) was used as a reference substance. No dilution
of the wood particles in KBr was required to obtain the spectra. Data analysis was
performed using WinFIRST sofiware from Thermo Nicolet (Madison, WI). No baseline

modification was done before performing data analysis of the spectra.

Based on prior work (1), the regions of interest in the FTIR spectra of modified
wood particles with maleated polyolefins were the absorbance bands near 2900 cm'1 and
1740 cm“, for CH stretching of aliphatic carbon chains and carbonyl group stretching
suggesting the formation of ester linkages, respectively. Using the integrated area under

these peaks, a grafting index (G1) was calculated using the following equation:

85

when

repre:

thein

Elect]
a non
10w 1
conce
OXF'Q‘
from
aIOms
hydrc
(nher
Garbo

carbo

11).

 

taktsl

 

A .
GI X _ X (Modified)

— (1)
Ax (Unmodified)

where x represents the absorbance band at either 2900 cm'l or 1740 cm", Ax (Modified)

represents the integrated area of the peak after modification and Ax (Unmodified) represents

the integrated peak area of the unmodified wood particles.

X-ray photoelectron spectroscopy (XPS) analysis was carried out on a Physical
Electronics Phi 5400 ESCA System, (Physical Electronics USA, Chanhassen, MN) using
a non-monochromatic Mg source and a takeoff angle of 45° relative to the detector. A
low resolution scan from 0 to 1100 eV binding energy was used to determine the
concentration of each element present on the surface of the samples, along with the
oxygen to carbon (O/C) atomic ratio, whereas a high resolution scan of the C1, region
from 280 to 300 eV was performed to further analyze the chemical bonding of the carbon
atoms. Carbon components C1 arise from carbon atoms bonded only to carbon and/or
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 non-
carbonyl 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 non-carbonyl group (O-C=O) (9, ll,
17). The procedure for XPS data collection and analysis was detailed in other articles (9,

11).

As mentioned, the chemical reaction of maleated polyolefins and wood particles

takes place between the maleate groups of the polyolefin and hydroxyl groups on the

86

surface of wood. Since C2 component arises from atoms bonded to a single oxygen
atom, other than a carbonyl oxygen (C-OH), the change in the content of C2 component
before and after modification can be used to monitor the occurrence of the esterification
between the maleated polyolefins and the wood particles. To quantify this change, a

hydroxyl index (HI) was calculated from C2 component of C1, data as follows:

 

C2 Unmodified

where CZModjficd and CZUnmodified represent C2 after modification and in the unmodified

wood particles, respectively.

A complimentary technique to FTIR and XPS analyses, titrimetric analysis was
also performed on unmodified and modified wood particles to provide additional proof of
the esterification reaction, and to elucidate the mechanism of the chemical reaction
between wood and MAPE. Titrations were carried out following a procedure described
elsewhere (17), using an Oakton pH CON 510 pH meter. Accurate endpoints were
difficult to determine when a visual indicator was used because wood particles
significantly darkened the solution color. To insure greater accuracy, potentiometric
titrations were used with a first derivative method for endpoint determination (18). Four
samples each of unmodified and modified wood particles were analyzed for acid value
(AV), saponification value (SV), and hydroxyl value (HV). Acid value accounts for the

amount of free carboxylic acid groups present in the sample, while saponification value

87

measures both acid and ester groups in the sample. Hydroxyl value quantifies the free
and accessible hydroxyl groups present in the sample (17). These quantities were

calculated using the following formulas (17):

AV = 56.1 1‘1— (3)
p
sv = 56.1(b--a)E (4)
p
HV = [(b — a)N (56.1) + AV (5)

where N is the normality of potassium hydroxide (KOH), p is the amount in grams of
wood particles, b is the volume of KOH needed to neutralize the blank (solvent only) and
a is the volume of KOH required to neutralize the sample. The factor 56.1 accounts for

the molecular weight of KOH.

88

RESULTS AND DISCUSSION

Effect of Maleated Polyethylene (MAPE) Content

FTIR spectroscopy was used to monitor and quantify changes that occurred on the
surface of wood particles after reactive extrusion with MAPE. Infrared spectra of
unmodified wood particles, pure MAPE, and wood particles modified with various
concentrations of MAPE are shown in Figure 4.3. Table 4.3 lists the wavenumbers of
peaks found in these spectra, along with assignments of corresponding functional groups.
Regardless of the MAPE content, evidence of the grafting of MAPE compound to the
surface of wood particles was apparent in the spectra of modified wood particles (spectra
C-F). A distinct change was clearly seen near the absorption band at 2900 cm", where a
large band between 2920 and 2850 cm", having two distinct peaks similar in appearance
to pure MAPE (spectrum B), replaced the single peak in the unmodified wood particle
(spectrum A). This change was most noticeable at MAPE concentrations above 10%
where the integrated area under the absorption band near 2900 cm'1 increased in the
modified wood particles, compared to the unmodified ones, implying the grafting of the
pendant polyethylene chain of MAPE to the surface of wood particles. Similarly, the
integrated area under the absorption band near 1740 cm'1 has significantly increased for
wood particles extruded with MAPE compound. Several authors have correlated the
increased integrated area or peak height of this band with the esterification reaction
between wood particles and maleated compounds, since absorbance in the range of 1725-

1750 cm'I is characteristic of ester carbonyl stretching (5-10, 12-14, 19).

89

 

93 2900
1740

8
1 6

1710
723

D
2919
285

In
3400

can—3c rang-mx—accx
-n o
3400
29
1740 1740
‘ 4
1122

 

 

 

4000 3500 3000 2500 2000 1 500 1000 500
Wavenumbers (cmA-1)

Figure 4.3. FTIR spectra of unmodified wood particles (spectrum A), pure maleated
polyethylene-MAPE (spectrum B), wood particles modified with 5%
MAPE (spectrum C), 10% MAPE (spectrum D), 15% MAPE (spectrum
E), and 20% MAPE (spectrum F) after a second 24-hour Soxhlet
extraction in the region4000-400 cm".

90

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Am A 5 ...............
2 SEE; moose a $8 ..... MS
a cone? wide :0 .......... 83% ..........
. . Smashed“. :o ”2 EN: . - ..... .
a 2 S 922% odd Ba 06 as: a: $2 8: 3.: mm: 82
2 maocafieoo £0 £2 ..... 8.2 ..... 82-8:
2 223553 £0 83.8: 83-3: 3: S: 23.3:
2.2 9.2an ma: 2582 as. _ do: as. _ do: .......... 8: -82
a a a a A—NAF—onao ”momv lllllllllll
o. 2 2 do msgaeaob a: E:
. . . . 2.8380 53.8 ........
o. 2 E d o messed Ono 8: 8: - - 2.:
2.2 9222:. 9:0 £635 .......... fl: WE ......
2.2.2 £0 .foao @222; :0 Sam; mom $3-83. 3088 33.2% 8%
2.3.2 3232:. :0 82.82 82.82 .......... 83-8mm
p.83 .552 9.53 $22 nae p-83 A as o8:
082%: ado—Swag 0.3.— 5? .358... a? .552... $22 $22 _.
35.55::
8.39::— coog 8.3:»:— vooB 9:..— 9:5
838E coo? BEBE
Ea .722 25 .mdfiz 8.5 gonad 8o? Homage: é acoaowama. 93 95m ooueomo< a: .3 2.3

91

The grafting efficiency of wood particles with various MAPE contents through a
reactive extrusion process was quantified by calculating the grafting index (Equation 1)
from the integrated area under the peaks near 2900 cm'1 and 1740 cm'1 for unmodified
and modified wood particles. Figure 4.4 clearly illustrates that the grafting index
increased with MAPE content up to 15%, independent of the absorption peak used.
There appeared to be some leveling off between 15% and 20% MAPE concentration,
which may indicate a level of maximum grafting efficiency has been reached. These
results suggest that the esterification reaction was a function of the MAPE concentration

used.

92

3.0

 

 

 

 

8 2.5

E

3

E 2.0

I:

:3.

S 1.5

U

w

E 1.0

g - -o--2900 cm"

x 0.5

< —C}—-1740cm"
0.0

 

0 5 10 15 20 25
MAPE Content (°/o)

Figure 4.4. Grafting index for FTIR absorbance bands near 2900 cm'1 and 1740 cm"
for unmodified wood particles and wood particles modified with 5-20%
MAPE.

93

XPS and titrimetric data listed in Tables 4.4 and 4.5, respectively, corroborated
the conclusions drawn from FTIR analysis. As expected, reactive extrusion of wood
particles with MAPE caused an increase in the concentration of unoxidized carbon atoms
(Cl component) and decreased the contents of oxidized carbon atoms (C2-C4
components). Consequently, a significant decrease of the O/C atomic ratios was
observed due to the presence of aliphatic carbons of polyethylene chains of MAPE.
Furthermore, both the hydroxyl index (HI in Table 4.4) and the free and accessible
hydroxyl groups (HV in Table 4.5) significantly decreased after modification with MAPE
compounds through a reactive extrusion process. The decrease in both hydroxyl index
and free and accessible hydroxyl groups after modification implied that esterification

reaction took place through the hydroxyl groups on the surface of the wood particles.

Unlike FTIR, which clearly showed an increase in ester carbonyl as a function of
MAPE content, C3 component (carbonyl groups) showed a decreasing trend as MAPE
content increased, while C4 component (ester groups) was not detected at all on the
surface of the modified wood particles. This apparent difference can be explained by the
higher surface sensitivity of XPS, which has a probing depth of only a few nanometers
(l l, 16). Unoxidized carbon from the pendant polyolefin chain of MAPE was
concentrated on the wood particle surface, as evidenced by the high content of C1
component (Table 4.4), and may have obstructed the detection of C3 and C4 components

during XPS analysis.

94

 

.N corms—om 80¢ toga—=28 22:: :9

.mD—n—Eflm 050m cm COHOOHDV mag “Tr—8&0 H—Omumowtuofio or: EOHM CNN

docombxo HoExom __ vm vacuum a cote 8:350 805 803 com: 8.253 uooB BEBE .28 35608:?

 

mod

36

_m.o

omd

26

co;

ME:
59:

IO

 

 

 

. . . . . . . . med: 2% a?
8 o 2 c 3” a we 3 8 o 8 o 8 8 8 N 8288 8328 825

. . . . . . . . max: 22 £3
8 o a _ om m ow No B o 3. o 3 mo 2 o 8285 8322 825

. . . . . . . . .232 $2 a?
8 c K m 8 2 NM 3 2 o E c an a a : 8288 3323 253

. . . . . . . . med: .2 as,
8 c we 4 cm 2 S a 2 o a o 2 mm 3. 2 B285 3323. Boa
85 2a Ks 2.2 :.o cod 5.2 8.2 $22 and
”no 2; cm. 2 m 3.3 So 85 m _ .3 a. _ m 8328 ES, 8288::
5 8 8 6 8:8 New u o

28on GE MECBNE
g 38o 5.6 $38.... 20 ..

mcocmmomfioo 3:082“;

 

max .3 358869 mac—omtam coo? mo 38¢ .20 :ouEOmoy—éwm: use muoEmanoU ooflBm Esau—m

den £an

95

Table 4.5. Hydroxyl Value, Acid Value and Saponification Value Determined by

 

 

 

 

Titrimetric Analysis
Titration values2
Wood particles' Hydroxyl value Acid value Saponification value
(HV) (AV) (SV)
(eq/kg) (eq/kg) (err/kg)
Unmodified 268 i 4 3.15 i 0.11 64 i 6
Modified
with 5% MAPE 233 j: 8 3.37 i 0.23 65 i 4
Modified
with10%MAPE 233i26 3.53i0.12 79i8
Modified
with15%MAPE 217111 3.87i0.20 132:10
Modified
with 20% MAPE 215:3 4.00:0.24 131:8

 

lUnmodified and modified wood particles used were those obtained after a second 24-h
Soxhlet extraction.

2Titration values represent an average of four samples.

96

As shown in Figure 4.1, two possible reactions could occur between MAPE and
wood particles in a reactive extrusion process: a single site reaction, which lead to the
formation of monoester with carboxylic acid pendent groups (Figure 4.1a) and/or or
diester formation without carboxylic acid pendent groups (Figure 4.1b) (9, 13). Acid
value (AV) and saponification value (SV) determined by titrimetric analysis (Table 5)
were used to elucidate which reaction had occurred. The AV, which accounts for the free
carboxylic acid group content in the system, slightly increased in the samples after
modification (Table 4.5). However, this increase was not significant compared to the
unmodified wood particles. Thus, there were a negligible amount of free carboxylic acid
groups in the system. By contrast, SV, which accounts for both ester and acid groups,
increased significantly after modification (Table 4.5). Since there was a negligible
amount of free carboxylic acid groups on the surface of the modified wood particles, the
increased SV originated mainly from the formation of maleate ester with OH groups on
the wood particle surface. This result suggests that most of the MAPE was attached to
the wood particle surface via two acid groups from the cyclic anhydride of the MAPE,
i.e., through diesterification reaction (Figure 4.1b). Moreover, FTIR spectra shown in
Figure 4.3 support this mechanism. Two distinct peaks in the range 1700-1750 cm'l
should have been detected if the esterification reaction had occurred through monoester
reaction (6, 9, 13). The first peak at around 1705-1710 cm'1 is associated with the non-
reacted carboxylic acid (monoester formation) whereas the second one near 1725-1750

cm'1

are caused by the ester carbonyl absorption of the reacted MAPE with the OH
groups of wood (diester formation) (6,9,13). As seen in Figure 4.3, all spectra of

modified wood particles showed only one peak at 1740 cm", which is characteristic of

97

ester carbonyl, and the absence of bands at 1705-1710 cm'1 in these spectra clearly
proved that there were no free carboxylic acid groups in the samples, thus confirming

diester formation.

The relatively high temperature (160°C) used to modify the wood particles in the
reactive extrusion process may also explain why the diester form was predominant in this
system. Other authors have reported higher monoester formation in the reaction of
maleated compounds with wood particles in solvents at room temperature (3-8).
Additional heat was required to cause the second acid group of the maleated compound to
react with wood particles, forming the diester (4,6). Hon and Xing determined that
diesters began to increase in their system at temperatures above 140°C (4). Similarly,
Felix and Gatenholm found more diester content than monoester in cellulosic particles
modified with MAPP at 100°C (13). Since the temperature used for reactive extrusion of
wood particles in this study was even higher than those reported by other authors, it was
likely sufficient to cause the reaction of both acid groups in the MAPE with hydroxyl

groups on the wood particle surface, resulting in diester as the primary ester form.

Although AV and SV were calculated, it was not possible to quantify the total
ester content in these samples because calculation of monoester, diester and total ester
content require sample weight gain after modification (3, 4). During the reactive
extrusion process, waste material is produced at the beginning and end of the run, so not
all modified sample can be collected. Even without this information, evidence of the
grafting reaction suggests that reactive extrusion is an efficient process for modifying

wood particles with binding agents such as MAPE compounds.

98

Effect of Extrusion Processing Conditions

In a prior part of this study, wood particles were modified at a constant extrusion
temperature profile (160°C) and rotational screw speed (60 rpm). Since the chemical
reaction between wood particles and maleated polyolefins can be affected by both
temperature and residence time in the extruder, it is important to understand how these
extrusion processing conditions would affect the grafting efficiency of wood particles

modified through reactive extrusion.

Figure 4.5 shows the grafting indices for the absorption bands near 2900 cm'1
(Figure 4.5a) and 1740 cm'1 (Figure 4.5b) plotted for each extruder barrel temperature
over the range of rotational screw speeds investigated. There was a clear indication that
the grafting indices increased after modification, irrespective of the extrusion temperature
profile and rotational screw speed used. The increased grafting index values indicated
that more CH2 and C=O groups had been attached to the wood particle surface, which
implied that the esterification reaction had occurred between the wood particles and
MAPE. However, surface modification was not a function of processing conditions used
since no trend in grafting index was observed with increasing the extruder’s temperature

profile or increasing rotational screw speed for either absorbance band.

99

AmMOdifiOd/Azgoounmodified

Amomodifiedlmmunmodified

Figure 4.5.

3.0

 

 

 

 

 

 

 

 

 

 

 

 

 

a
2.5 ( )
1‘21
2.0 X . —-e--130°C
o ---- 9 """ V ——r:}— 140°C

1.5 ---A--- 160°C
- -o - —180°C

1.0 X )l( X $K +Control

0.5

0.0 .

O 20 40 60 80 100
Rotational Screw Speed (rpm)

3.0

2.5

2.0 mew-130°C
—D— 140°C

1.5 ----A--160°C
- -e - —180°C

1.0 X )K >K X +Control

0.5

0.0

0 20 40 60 80 100

Rotational Screw Speed (rpm)

Effect of processing conditions on grafting index for unmodified and
modified wood particles over the range of extruder barrel temperature and
rotational screw speed combinations: (a) band near 2900 cm'] and (b) band
near 1740 cm".

100

XPS results agreed with the above conclusions obtained from the infrared
spectroscopy (Tables 4.6 and 4.7). Changing the extruder’s rotational screw speeds while
maintaining a constant temperature profile (Table 4.6) or vice versa (Table 4.7) did not
affect the grafting efficiency. In both cases, it was apparent that the surfaces of modified
wood particles had been chemically changed from the unmodified wood particles, as
evidenced by the large increase in the concentration of unoxidized carbon atoms (C1
component) coupled with a significant reduction in both the O/C atomic ratios and
hydroxyl indices (HI) under all processing conditions. However, no trend was observed
in C1, H1 or the O/C atomic ratio as the extruder’s rotational screw speed or barrel

temperature was increased.

101

 

 

 

 

 

 

 

 

 

 

 

Table 4.6. Effect of Extruder’s Rotational Screw Speed on Surface Chemistry of
Wood Particles Modified with 20% MAPE at 160°C
, Elemental
$821231]; compositions O/C Analysis of C1, peaks (%) 0H
(%) atomic index
screw speed ratios (HI)
(rpm) 0 C c1 C2 C3 C4
”nexm‘ded 31.85 68.15 0.47 39.75 51.20 8.46 0.58 1.00
(control)
20 5.61 94.39 0.06 92.62 5.12 2.26 0.00 0.10
40 3.72 96.28 0.04 95.80 3.40 0.80 0.00 0.07
60 2.39 97.61 0.02 96.90 3.10 0.00 0.00 0.06
80 2.97 97.03 0.03 95.08 4.34 0.58 0.00 0.08
Table 4.7. Effect of Extruder’s Barrel Temperature on Surface Chemistry of Wood
Particles Modified with 20% MAPE at 60 rpm
, Elemental
E123: s compositions O/C Analysis of C1, peaks (%) OH
tern r atur e (%) atomic index
pe ratios (HI)
(°C) 0 C C1 C2 C3 C4
unem’ded 31.85 68.15 0.47 39.75 51.20 8.46 0.58 1.00
(control)
130 3.36 96.64 0.03 95.27 3.04 1.69 0.00 0.06
140 5.47 94.53 0.06 92.61 5.26 2.12 0.00 0.10
160 2.39 97.61 0.02 96.90 3.10 0.00 0.00 0.06
180 5.05 94.95 0.05 95.24 4.76 0.00 0.00 0.09

Effect of Maleated Polypropylene (MAPP) Molecular Weight

Maleated polyolefins with a wide variety of molecular weights are available on
the market. Four different MAPP compounds were used (Table 4.1) to determine how
different molecular weights (from a low of 11,200 to a high of 52,000 g/mol) affect the
surface modification of wood particles in reactive extrusion process. To achieve this
objective, the MAPP content was held constant at 20% while the extrusion temperature

profile and rotational speed of the screws were set at 160°C and 60 rpm, respectively.

Figures 4.6 and 4.7 illustrate the FTIR spectra of pure MAPP compounds with
various molecular weights and wood particles modified with these MAPPs, respectively.
The peak assignments for these spectra are also listed in Table 4.3. Evidence of the
grafting of MAPP to the wood particles through esterification is apparent in the modified
wood spectra (Figure 4.7, spectra B-E) because the single absorption band near 2900 cm'1
in the spectrum of unmodified wood particles has been replaced by a large band between
2955 and 2840 cm", having distinct peaks similar in appearance to the pure MAPP
counterparts. Another indication of grafting of the maleated polyolefin was an increased
area of the absorption band at around 1740 cm", from C=O groups attached to the wood
surface, likely from the esterification reaction (5-10, 12-14) Although not clearly seen in
Figure 4.7 due to stacking of five spectra in one figure, the individual spectra of modified
wood particles showed an increased intensity in the bands at 1462 and 1376 cm'1 which
also indicated the grafting reaction had occurred and added more C-H to the surface of

the modified samples (10, 12, 14).

103

 

xacg-mx—oacx

 

“was:

 

 

 

A

4000 3500 3000 2500 2000 1500 1000 500
Wavenumbers (cmA-1)

 

Figure 4.6. FTIR spectra of pure MAPP compounds with various molecular weights:
11,200 g/mol or E-43 (spectrum A), 39,000 g/mol or G-32l6 (spectrum
B), 47,000 g/mol or G-3015 (spectrum C) and 52,000 g/mol or G-3003
(spectrum D) in the region 4000-400 cm".

104

 

O
O
O)
N
(I)
V-
05
N

racgonr—wacx

 

to ur-mla C:

 

 

 

f

4000 3500 3000 2500 2000 1500 1000 500
Wavenumbers (cm"-1)

Figure 4.7 FTIR spectra of unmodified wood particles (spectrum A), wood particles
modified with MAPP compounds of various molecular weights: 11,200
g/mol or E-43 (spectrum B), 39,000 g/mol or G-3216 (spectrum C),
47,000 g/mol or G-3015 (spectrum D), and 52,000 g/mol or G-3003
(spectrum E) in the region 4000-400 cm' .

105

Moreover, all four MAPP compounds had significantly increased grafting indices
as compared to unmodified wood particles (Figure 4.8). However, no significant
difference in grafting index was observed between the four types of MAPP-modified
wood particles. These results were in agreement with those obtained from the XPS
analysis (Table 4.8). As expected, the surface of modified wood particles resulted in an
increased CI content and a significant decrease in both O/C atomic ratio and hydroxyl
index (HI), all of which provided evidence of the surface modification through

esterification reaction.

106

3.0

 

 

 

 

 

'3 2.5

E

3 2 0

E .

C

:3.

5 1.5

8

E 10

‘3 ' -—.<>-—2900 cm"

E

4’? 0,5 —1:1——1740cm'1
0.0

0 11,200 39,000 47,000 52,000

Mw of MAPP (glmol)

Figure 4.8. Effect of molecular weight on grafting index for unmodified wood
particles and wood particles modified with 20% MAPP compounds.

107

 

Table 4.8. Elemental Surface Compositions and High-Resolution C15 Peaks of
MAPP, Wood Particles, and Wood Particles Modified with MAPP

 

 

 

 

 

Determined by XPS
Elemental
compositions O/C Analysis of Cl, peaks (%) OH
Materials (%) atomic index
0 C mm C1 C2 C3 C4 (m)
5'43 4 89 95 11 0 05 94 92 4 35 0 74 0 0 O8
(Mw=11,200) ‘ ' ‘ ‘ ' ' ° '
G-3216
(Mw=39,000) 3.86 96.14 0.04 95.85 3.24 0.91 O 0.06
G-3015
(Mw=47,000) 3.77 96.23 0.04 96.01 2.62 1.37 O 0.05
G-3003
(Mw=52,000) 3.28 96.72 0.03 96.58 2.57 0.84 0 0.05
Unmodified Wood 31.85 68.15 0.47 39.75 51.20 8.46 0.58 1.00
Wood-E-43 7.22 92.78 0.08 93.39 4.75 1.86 0 0.09
Wood-03216 3.60 96.40 0.04 94.86 4.17 0.97 0 0.08
Wood-G-3015 3.36 96.64 0.04 95.38 4.20 0.42 O 0.08
Wood-G-3003 5.87 94.13 0.06 90.99 6.19 2.83 0 0.12

 

108

However, no distinct trend was observed between molecular weight of MAPP and
grafting efficiency through a reactive extrusion process. Since both FTIR and XPS
results confirmed that there were no significant differences in grafting of MAPP with
various molecular weights, the molecular weight range studied (11,200-52,000 g/mol)
may not have been large enough to observe differences in reactivity. Another possibility
could be that at 20% MAPP content, the maximum level of grafting had already been
achieved with each compound, regardless of its molecular weight. Thus the differences
in the grafting efficiency were not detected. Moreover, Figure 4.4 clearly illustrates that
the GI leveled off between 15 and 20% maleated polyolefin content, which may indicate
a level of maximum grafting efficiency had been reached. Perhaps another study at lower
MAPP concentrations would elucidate differences of MAPP molecular weight on the

efficiency of the grafting reaction.

109

CONCLUSIONS

This study examined the chemical reactions between maleated polyolefins and
wood particles in a reactive extrusion process. The effects of maleated MAPE content,
extrusion processing conditions (barrel temperature and rotational screw speed), and
molecular weight of MAPP were studied, with the goal of determining the effects of each
set of conditions on the grafting efficiency of modified wood particles in a reactive
extrusion process. Efficiency of the modification was evaluated using FTIR, XPS and
titrimetric analysis. From the experimental results, the following conclusions can be

drawn:

1. The esterification reaction between wood particles and MAPE was a function of
the MAPE concentration used to modify the wood particles. The grafting reaction
produced mostly the diester form of the modified wood particle during reactive

extrusion.

2. No significant difference was found in grafting efficiency of the modified wood
particles at different extrusion processing conditions. Changing the extruder’s
barrel temperature profile (130-180°C) and/or its rotational screw speed (20-80
rprn) resulted in adequate grafting, which indicated that the esterification reaction

was not a function of processing conditions over the range studied.

3. Regardless of MAPP molecular weight (from a low of 11,200 to a high of 52,000
g/mol) all investigated MAPP compounds were effective in changing the surface
of wood particles after modification, compared to unmodified wood particles.
However, no distinct trend was observed between molecular weight of MAPP and

grafting efficiency through a reactive extrusion process. Iit is believed that the

110

high content of MAPP (20%) used in this study prevented the detection of
differences in the grafting efficiency because the maximum level of grafting

reaction had already occurred with each component at 20% MAPP content.

. The reactive extrusion process was found to be a suitable way to modify wood
particles with maleated polyolefins as it worked quickly and without the use of
solvents. This process would be industrially friendly, and in light of our prior
success bonding the modified fibers together in a hot press, would allow large
quantities of modified wood particles to be produced for the manufacture of a

formaldehyde-free wood composite product.

111

10.

11.

REFERENCES

Carlbom, K. and Matuana, L.M., "Composite materials manufactured from wood
particles modified through a reactive extrusion process," Polymer Composites, 26
(4): 534-541 (2005).

Matuana, L.M. and Carlbom, K., US Provisional Patent # 60/592,918, July 30,
(2004)

Matsuda, H., "Preparation and utilization of esterified woods bearing carboxyl
groups," Wood Science and Technology, 21 (1): 75-88 (1987).

Hon, D.N.S. and Xing, L.M., "Thermoplasticization of wood. Esterification," in
Viscoelasticity of Biomaterials, W.G. Glasser and H. Hatakeyama, Editors.
American Chemical Society: Washington DC. p. 118-132 (1992).

Marcovich, N.E., Reboredo, M.M., and Arangguren, M.I., "Sawdust modification.
Maleic anhydride chemical treatment," Holz als Roh- und Werkstoff, 54 (3): 189-
193(1996)

Aranguren, M.I., Marcovich, N.E., and Reboredo, M.M. "Sawdust and woodflour:
Its esterification and use in the formulation of polymer composites," in Recent
Advances in Biotechnology for Tree Conservation and Management, Proceedings

of an IFS Workshop, Brazil, (1998).

Timar, M.C., Mihai, M.D., Maher, K., and Irle, M., "Preparation of wood with
thermoplastic properties. Part 1. Classical synthesis," Holzforschung, 54 (l): 71-
76 (2000).

Timar, M.C., Maher, K., Irle, M., and Mihai, M.D., "Preparation of wood with
thermoplastic properties. Part 2. Simplified technologies," Holzforschung, 54 (1):
77-82 (2000).

Matuana, L.M., Balatinecz, J .J ., Sodhi, R.N.S., and Park, C.B., "Surface
characterization of esterified cellulosic fibers by XPS and F TIR Spectroscopy,"
Wood Science and Technology, 35 (3): 191-201 (2001).

Kazayawoko, M., Balatinecz, J .J ., and Woodhams, R.T., "Diffuse reflectance
Fourier transform infrared spectra of wood fibers treated with maleated
polypropylenes," Journal of Applied Polymer Science, 66 (6): 1163-1173 (1997).

Kazayawoko, M., Balatinecz, J .J ., and Sodhi, R.N.S., "X-ray photoelectron

spectroscopy of maleated polypropylene-treated wood fibers in a high-intensity
thermokinetic mixer," Wood Science and Technology, 33 (5): 359-3 72 (1999).

112

12.

13.

14.

15.

16.

17.

l8.

l9.

Kazayawoko, M., Balatinecz, J.J., and Matuana, L.M., "Surface modification and

adhesion mechanisms in wood fiber-polypropylene composites," Journal of
Materials Science, 34 (24): 6189-6199 (1999).

Felix, J .M. and Gatenholm, P., "The nature of adhesion in composites of modified
cellulose fibers and polypropylene," Journal of Applied Polymer Science, 42 (3):
609-20 (1991).

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

Cheremisinoff, N.P., Guidebook to Extrusion Technology. Prentice Hall:
Englewood Cliffs, NJ. (1993).

Ratner, 8D. and Castner, D.G., "Electron Spectroscopy for Chemical Analysis,"
in Surface Analysis .' The Principal Techniques, J .C. Vickerman, Editor. John
Wiley & Sons: Chichester, England. p. 43-98 (1997).

Anonymous, "Recommended methods for the analysis of alkyd resins," Pure and
Applied Chemistry, 33 (2-3): 411-435 (1973).

Harris, D.C., Quantitative chemical analysis, 4th Ed., W.H. Freeman: New York.
Chapter 12 (1995).

Socrates, G., Infrared characteristic group frequencies: tables and charts, 2nd
Ed., John Wiley & Sons, Inc.: New York (1994).

113

CHAPTER 5

INFLUENCE OF PROCESSING CONDITIONS AND MATERIAL
COMPOSITIONS ON THE PERFORMANCE OF F ORMALDEHYDE-F REE

WOOD-BASED COMPOSITES

This chapter is in press for Polymer Composites (2006). It is co-authored by K. Carlbom
and L.M. Matuana.

114

ABSTRACT

This study examined the differences between formaldehyde-free wood composite
panels made with maleated polyethylene (MAPE) and maleated polypropylene (MAPP)
binding agents. Specifically, the study investigated the contrasts of (i) base resin type,
PE vs. PP, (ii) molecular weight/maleic anhydride content in MAPP binding agents, and
(iii) the manufacturing methods (reactive extrusion vs. hot press) on the physico-
mechanical properties of the composites. FTIR and XPS analyses of unmodified and
modified wood particles after reactive extrusion with maleated polyolefins provided
evidence of chemical bonding between the hydroxyl groups of wood particles and
maleated polyolefins. While extruding the particles before panel pressing gave better
internal bond strength, superior bending properties were obtained through compression
molding alone. MAPP-based panels outperformed MAPE-based panels in stiffness.
Conversely, MAPE increased the IB strength of the panels compared to MAPP. Polymer
base resin had no effect on modulus of rupture or screw holding capacity. Differences
between the two maleated polypropylene compounds were not significant for any of the
mechanical properties tested. Formaldehyde-free wood composites manufactured in this
study often outperformed standard requirements for conventional particleboard,

regardless of material composition or manufacturing method used.

115

INTRODUCTION

The chemical reaction between wood and anhydride compounds has been shown
to plasticize the wood, making it more versatile for wood composite applications.
Several investigators have studied wood modification with anhydrides in order to
determine thermal properties, moisture resistance, compatibility with polymeric matrices,
etc. (l-ll). A lower softening temperature, compared to unmodified wood, has been
documented in esterified wood after reaction with various anhydrides (2, 8, 10, ll).
Esterified wood has also been found to resist moisture (l, 3, 4-11) and to be more
compatible than unmodified wood with polyester matrices in the manufacture of
composites (4-6). A second reaction of the esterified wood with epoxides formed
oligoesters, which further thermoplasticized the wood and allowed for better control of

structure and resulting properties of the modified wood (1, 7-11).

While the thermoplastic nature of esterified and oligoesterified wood modified
through these methods has been documented, few authors have exploited this
characteristic to make wood composites without added adhesive. Matsuda and co-
workers used three dicarboxylic acid anhydrides (maleic, phthalic and succinic
anhydride) to esterify wood particles, which were then molded into sheets through
compression molding ( 1, 8). Clemons and co-workers used similar methods to modify
wood fibers with anhydrides to prepare fiberboards (3). In both studies, chemically
modified wood was able to partially melt when heated and bond without additional
adhesive. Formulations containing succinic anhydride were found to have the most
thermoplastic character in both studies (1 , 3, 8). Further work by Matsuda et a1. (1 , 7, 8)

showed enhanced thermoplasticization of wood through the grafting of various types of

116

epoxides onto the already esterified wood surface. The oligoesterified wood produced
from these reactions was even more thermoplastic-like in structure than the esterified
wood produced through reaction with the anhydrides, and could be easily molded. Using
various types of epoxides allowed for different chemical structures in the
thermoplasticized wood, and could also be used to form crosslinked wood composites
which resembled plastic more than wood (1, 7, 8). Recently, Timar and co-workers
demonstrated the use of maleic anhydride to esterify wood particles, followed by
oligoesterification with glycidyl methacrylate and additional maleic anhydride as a
process to form thermoplastic-like wood particles (9-11). Panels formed through the
compression molding of modified wood particles displayed mechanical properties that
met or exceeded standard requirements for bending strength and internal bond strength of
particleboard or fiberboard (11). These wood composites were also found to be resistant

to fungal decay.

Although esterifying and oligoesterifying the wood surface has been shown to be
an effective method for manufacturing wood composite materials without additional
adhesive, most methods require harmful solvents such as xylene, dimethyl sulfoxide, or
N,N-dimethylformamide as part of the modification process (1-6, 8-11). Additionally,
the oligoesterification reactions entail a two-step process (or more depending on the
precise control of chemical structure desired) to functionalize the wood surface, which is
both time consuming and complicated (1, 7-11). Removal of solvents and further drying

of the wood particles is required before compression molding into panels (1, 3, 8, 11).

A simpler approach of functionalizing wood particles to produce wood composite

panels without additional adhesive has been developed in our prior work, which

117

demonstrated the concept of using a reactive extrusion process as a means of developing
a new, formaldehyde-free binding system for wood composite products (12). It has been
shown that the surface modification of wood particles with maleated polyolefins as the
binding agent could be achieved through a continuous reactive extrusion process without
the use of solvents (i.e., dry process). Surface analysis of the modified wood particles
indicated that the maleate groups of the binding agent were attached to the surface of the
wood particles via an esterification reaction, exposing polyolefin chains on the surface of
the particles (12, 13). The modified wood particles were then compression molded to
produce formaldehyde-free wood composite panels. During panel manufacturing, the
pendant polyolefin chains attached to the wood particles melted and flowed under heat
and pressure in the hot press, forming entanglements. The entangled polymer chains
locked together after cooling, forming direct particle-particle bonds. The physico-
mechanical properties of these composite panels, which contain no formaldehyde-based
adhesive, compared favorably with or even exceed current standard requirements for

particleboard of medium density.

Although formaldehyde-free wood composite panels were successfully
manufactured through a reactive extrusion process (12), the process required two
different steps: (i) reaction of the wood particles with the maleated polyolefin, and (ii) hot
pressing the modified particles into panels. A method that could reduce this process to
one step would save time and reduce production costs. Therefore, we hypothesized that
both the esterification reaction and binding into panels could occur simultaneously during

compression molding, which would greatly simplify the panel manufacturing process.

118

Furthermore, there currently are many types of maleated polyolefins, differing in
molecular weight, base resin type, acid value, percentage of maleic anhydride, etc.
Research has not been carried out to examine the effects of these maleated polyolefin

variables on the performance of formaldehyde-free wood composites.

Thus, the objective of this work was to investigate the effects of material
compositions as (i) base resin type of the maleated polyolefin, either polyethylene or
polypropylene, and (ii) maleic anhydride content/molecular weight of maleated
polypropylene, as well as manufacturing methods (reactive extrusion vs. hot press) on the
physico-mechanical properties of the composites. Emphasis was also placed on
determining the effectiveness of the modification after reactive extrusion of wood
particles with maleated polyolefins using Fourier transform infrared (FTIR) spectroscopy

and X-ray photoelectron spectroscopy (XPS).

119

EXPERIMENTAL
Materials

Maple wood particles of 425 micron (40-mesh) and 150 micron (100-mesh) size
were supplied by American Wood Fibers (Schofield, WI) and were used as particles. The
smaller particles (150 micron) were used for surface characterization to minimize the
effects of scattering and specular reflectance in the diffuse reflectance IR analysis.
However, panel manufacturing required a large quantity of modified particles, so larger
(425 micron) particles, which were easier to process, were used in panel manufacturing
and property testing. Hydrated zinc acetate, the catalyst, and xylene (99.9%, ACS
Grade), the solvent used for Soxhlet extraction, were obtained from Baker Analytical
Reagents (JT Baker Co., Phillipsburg, NJ). Maleated polyethylene-MAPE (G-2608) and
two maleated polypropylenes-MAPP (G-3003 and G-3015) supplied by Eastman
Chemical Co. (Kingsport, TN) were used as binding agents. Characteristics of the

maleated compounds are listed in Table 5.1. All chemicals were used as received.

120

Table 5.1. Characteristics of the maleated polyolefins used as binding agents

 

 

Properties MAPE MAPP MAPP
G-2608 G-3003 G-3015
Weight % maleic anhydride 1.5 1.5 2.5
Melting point (°C) 122 156 155
Average molecular weight (g/mol) 51,700 52,000 47,000
lMelt flow index at 190°C 8 12.7
Viscosity at 190°C --- 60,000 25,000

 

lMelt flow index measured at 190°C and 2.16 kg according to ASTM D1238.

121

Surface Characterization of Wood Particles

For batch compounding, 150 micron size wood particles were dried for 48 hours
at 105°C to a final moisture content of less than one percent before processing. A 10-1iter
high intensity mixer (Papenmeier TGAHK20) was used for dry blending of the wood
particles, binding agent, and catalyst. The woodzbinding agentzcatalyst weight ratio was
79:20:l. All components were combined in the mixer and blended for 10 minutes at

room temperature.

Reactive extrusion of wood particles was achieved by feeding the compounded
wood particles into a 32 mm conical counter rotating twin-screw extruder (C. W.
Brabender Instruments, Inc.) with an L/D ratio of 13:1, driven by a 7.5 hp lntelli-Torque
Plasti-Corder Torque Rheometer®. The barrel temperatures for the three zones inside the
extruder were set at 160°C for maleated polyethylene and 165°C for maleated
polypropylene, and the rotational speed of the screws was held at 60 rpm during the

experiments.

Following the procedures used in prior work, unmodified wood particles and
wood particles that had been modified through reactive extrusion were Soxhlet extracted,
dried to a constant weight at 105°C and analyzed by FTIR (12, 13). Based on previous
studies (12, 13), the regions of interest in the FTIR spectra of wood particles modified
with maleated polyolefins were the absorbance bands near 2900 cm'1 and 1740 cm", for
CH stretching of aliphatic carbon chains and carbonyl group stretching suggesting the
formation of ester linkages, respectively. Using the integrated area under these peaks, 3

grafting index (G1) was calculated using the following equation:

122

A x (Modified)
Ax (Unmodified)

 

GIx = (1)

where x represents the absorbance band at either 2900 cm'I or 1740 cm", Ax (Modified)

represents the integrated area of the peak after modification and AK (Unmodified) represents

the integrated peak area of the unmodified wood particles (13).

X-ray photoelectron spectroscopy (XPS) was performed on both unmodified
wood and wood modified with maleated polyolefin binding agents following the
procedure used in prior work (12, 13). A high resolution scan of the C1, region from 280
to 300 eV was run to elucidate the chemical bonding of the carbon atoms. Curve fitting
of this high resolution area resolved four different forms of carbon for unmodified wood
particles, whereas three different forms were resolved for modified wood particles.
Carbon components C1 arise from carbon atoms bonded only to carbon and/or 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 non-carbonyl 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 non-carbonyl group (O-C=O). Particular attention
was given to the C2 component from the high resolution scan because C2 component
arises from atoms bonded to a single oxygen atom, other than a carbonyl oxygen (C-OH).
Therefore, change in the content of C2 component before and after modification can be
used to monitor the occurrence of the esterification between the maleated polyolefins and

the wood particles, which occurs at the hydroxyl groups on the wood surface. To

123

quantify this change, hydroxyl index (HI) was calculated from C2 component of C], data

as follows:

C2 Unmodified

where C2M0dified and CZUnmodified represent C2 after modification and in the unmodified

wood particles, respectively (13).

Panel Manufacture

Larger wood particles (425 micron size) were compounded at room temperature

using the same procedure described above for the surface analysis.

Following the mixing step, two different methods were used to manufacture the

composite panels from the compounded wood particles as follows:

The first method was a one-step process where the compounded wood particle
mixtures were directly hot pressed without the reactive extrusion step. Compression
molding was performed using a hydraulic press from Erie Mill Co. (Erie, PA). Panels
were pressed at 193°C for 7 minutes using 8 MPa pressure. After pressing, panels were
removed from the press and cooled at room temperature under compression for 15
minutes. Panel dimensions were 380 by 380 by 6 mm, with a target density of 720

kg/m3 .

124

The second method was a two-step process where wood particles were modified
with maleated polyolefins in a reactive extrusion process, and then compression molded
in a hot press. Reactive extrusion of wood particles followed the same procedure as
previously described for the modification of wood particles for surface characterization.
Once extruded, the wood particles were compression molded into panels using the above-

described pressing conditions.

Panel Property Testing

Density was measured by two different methods for all panels: (i) a simple mass
over volume calculation for three panels of each type and (ii) internal density profile (X-
ray density analysis) using a Quintek QMS Density Profiler, model QDP-OIX, with 5

replicates per panel type.

Three-point flexural, internal bond (IB) strength and screw holding capacity tests
were performed on an Instron 4206 testing machine (using Series IX software) in
accordance with procedures outlined in ASTM standard D1037-99 (14). The crosshead
speeds were 3.05 mm/min, 8.13 mm/min, and 0.6 mm/min for flexural, 1B, and screw
holding capacity tests, respectively. Screw holding capacity was carried out from the
face of the panels. At least six samples were tested to obtain an average value for
modulus of rupture (MOR), modulus of elasticity (MOE), IB strength and screw holding
capacity, all of which were compared with values listed for particleboard of medium

density in the standard ANSI A208.1-l999 (15).

125

Statistical Analysis

A two-sample t-test was carried out with an or significance value of 0.05 to
determine the effects of material compositions and manufacturing method on the density,
flexural, internal bond and screw holding properties of the composites. Comparisons
between binding agents’ base resin types and maleic anhydride contents/molecular
weights were made under one manufacturing method. Whereas comparisons between
manufacturing methods were performed under one base resin type or maleic anhydride
content/molecular weight. All statistical analysis was performed using Design Expert

software (Version 6) from Stat-Ease, Inc. Minneapolis, MN.

126

RESULTS AND DISCUSSION
Surface Characterization of Wood Particles

As mentioned, the reaction of wood particles with maleated polyolefins occurs
between maleate groups of the maleated polyolefin and hydroxyl groups on the surface of
wood, which lead to the formation of a monoester with carboxylic acid pendant groups
(Figure 5.1a) and/or diester formation where both carboxylic acid groups react with the
wood particle surface (Figure 5.1b). In order to form a wood composite panel, the
pendant polyolefin chains grafted to the wood surfaces must melt and flow under heat
and pressure during compression molding, forming entanglements. Upon cooling, these

entangled polymer chains would be locked together, forming the panel (Figure 5.2).

127

11
0-0—CH2

/
o / H 0’9 Monoester
/ a) '
OH CH, o
/ + 0 |
CH or
; OH >/ M \ ('3'
O b) / o—c—-CH2
/
Wood / O—C—C
part1cle Maleated / 8 H Di-ester
polyolefin
Modified wood
particle
Figure 5.1. Modification scheme for esterification reaction between wood particles

and maleated polyolefins: (a) monoester and (b) diester formation.

 

 

 

 

o 0
II II
/ l—O—C—CH2 + HZC—C—O— \
/ \
/ _O_C__S/\/\/\ Wfi—fi-O“§
/ ('5 o
h
Modified wood particle l 2 Modified wood particle
I

 

 

 

Figure 5.2. Panel manufacturing scheme for the two-step method showing the
bonding of pre-reacted wood particles.

128

F TIR spectroscopy was used to observe and quantify changes that occurred on the
surfaces of wood particles after reactive extrusion with maleated polyolefins. FTIR
spectra of unmodified wood, maleated polyethylene (MAPE G-2608) and wood particles
modified with MAPE after reactive extrusion are shown in Figure 5.3. As stated earlier,
the bands of interest in the FTIR spectra occurred near 2900 cm'1 and 1740 cm". In the
spectrum of modified wood particles (spectrum C), two distinct peaks between 2927 cm'l
and 2853 cm'1 similar in appearance to pure MAPE (spectrum B) replaced the single peak
near 2900 cm'1 in the unmodified wood (spectrum A). The integrated area under this
band increased after modification with MAPE (spectrum C), implying the grafting of the
polyolefin chain of MAPE to the surface of the wood particles. Similarly, the integrated
area under the absorption band near 1740 cm" significantly increased for wood particles
extruded with MAPE. Several authors have correlated the increased integrated area or
peak height of this band with the esterification reaction between wood particles and
maleated compounds, since absorbance in the range of 1725-1750 cm'1 is characteristic of
ester carbonyl stretching (1-6, 9, 10). Although not shown, similar infrared spectra were
obtained for wood particles treated with maleated polypropylene binding agents, i.e.,

MAPP G-3003 and G-3015.

129

 

 

1710

racguur—aacx

mama:

 

 

 

 

4000 3500 3000 2500 2000 1500 1000 500
Wavenu m bers (cm *-1)

Figure 5.3. Example FTIR spectra of unmodified wood particles (A), pure MAPEG-
2608 (B), and wood particles modified with MAPEG-2608 (C) in the
region 4000 to 400 cm".

130

The grafting efficiency of wood particles modified with MAPE (G-2608) and two
types of MAPP (G-3003 and G-3015) through a reactive extrusion process was quantified
by calculating the grafting index (Equation 1) from the integrated area under the peaks
near 2900 cm'1 and 1740 cm'1 for unmodified and modified wood particles. Results
listed in Table 5.2 show that grafting index increased after reactive extrusion with each
maleated polyolefin, independent of the absorption peak used. These results suggest the

esterification reaction had occurred during reactive extrusion.

131

Table 5.2. Grafting index for peaks near 2900 cm'1 and 1740 cm'1 for unmodified and
modified wood particles with various maleated polyolefin compounds

 

 

 

 

 

Grafting Index (GI)l
Samples 1 1
2900 cm' 1740 cm'
Unmodified wood particles 1.0 1.0
Wood modified with MAPE G-2608 2.3 2.1
Wood modified with MAPP G-3003 1.9 1.9
Wood modified with MAPP G-3015 1.8 1.9

 

1G1 is calculated using Equation 1.

132

XPS data listed in Table 5.3 supported the results of FTIR analysis. As expected,
reactive extrusion of wood particles with each of the maleated polyolefins caused a
significant increase in the concentration of unoxidized carbon atoms (Cl component) and
corresponding decrease in the contents of oxidized carbon atoms (C2-C4 components) in
the modified wood particles. Additionally, the hydroxyl index (HI) has significantly
decreased after modification with MAPE or MAPP compounds through a reactive
extrusion process. The reduction in hydroxyl index suggested that for all three maleated
polyolefins, the esterification reaction took place through the hydroxyl groups on the
surface of the wood particles.

While the grafting index calculated from the F TIR analysis for the band near 1740
cm'1 indicated an increase in ester groups after modification, regardless of the type of
maleated polyolefin used, C3 component (carbonyl groups) was reduced in all samples
after reactive extrusion with maleated polyolefins, while C4 component (ester groups)
was not detected at all on the surface of the modified wood particles (Table 5.3). This
apparent difference can be explained by the higher surface sensitivity of XPS, which has
a probing depth of only a few nanometers (l6). Unoxidized carbon from the pendant
polyolefin chain of MAPE or MAPP was concentrated on the wood particle surface, as
evidenced by the high content of C1 component (Table 5.3), and may have obstructed the

detection of C3 and C4 components during XPS analysis (13).

133

Table 5.3. High-Resolution C15 Peaks of Wood Particles Determined by XPS

 

 

 

 

 

Analysis of C15 Peaks (%) 011 index
Materials
C1 C2 C3 C4 (I-II)‘
uandlfied W°°d 39.75 51 .20 8.46 0.58 1,00
partrcles
Pure MAPE G-2608
(Mw=51,700 g/mol) 87.13 7.71 5.16 0.00 0,15
Wood modified with
MAPE G-2608 95.08 4.34 0.58 0.00 0.08
Pure MAPP G-3003
(Mw=52,000 g/mol) 96.58 2.57 0.84 0.00 005
Wood modified with
MAPP G-3003 90.99 6.19 2.83 0.00 0.12
Pure MAPP G-3015
(sz47’000 g/mol) 96.01 2.62 1.37 0.00 005
Wood modified with
MAPP @3015 95.38 4.20 0.42 0.00 0,08

 

lOH index calculated from Equation 2.

134

Physico-Mechanical Properties

Density

Table 5.4 summarizes the density (calculated and profile) for the panels
manufactured in this study. An example of the x-ray density profile, illustrating the

density of the sample at the face and core regions, is shown in Figure 5.4.

Experimental panels were within the medium density range as specified in the
ANSI standard A208.1, with average calculated density values ranging from 775-780
kg/m3 (Table 5.4). The calculated and overall (x-ray profile) density of the panels was
nearly the same, regardless of the panel manufacturing method. However, the
manufacturing method showed two distinct trends in the density profiles of the composite
panels. Panels made from unextruded wood particles (one step process, i.e. hot press
only) had a higher density in the face region than those containing extruded wood
particles (two step process, e.g., reactive extrusion followed by hot press). This feature is
typical of conventional glued particleboard, due to greater compaction and solidification
of the faces of the panels which experience direct heat during the pressing cycle (17, 18).
Conversely, panels manufactured from extruded wood particles had a higher density in

the core region.

135

 

Table 5.4. Density data for experimental panels bound with maleated polyolefins

 

X-ray Density Profile

 

 

 

 

Calculated
Panel Types Densi Overall Face Core
(kg/m (kg/m5 (kg/m3) (kg/m3)
MAPE G-2608 — unextruded 780 780 905 748
MAPE G-2608 — extruded 778 779 881 759
MAPP G-3015 — unextruded 780 781 906 745
MAPP G-3015 - extruded 782 783 833 783
MAPP G-3003 — unextruded 775 778 866 760
MAPP G-3003 — extruded 775 776 821 760

 

1000 ——-— - -+—~ -» - ------

 

 

 

 

 

 

 

 

 

g ’\ /

. Core
5 600 Face Face
:E‘
8 400
0)
o

200

o ,
0 1 2 3 4 5 6

Thickness (mm)

Figure 5.4. X-ray density profile of a sample made from unextruded MAPP G-3003,
illustrating the face and core regions of a typical sample.

136

Mechanical properties of a material are strongly influenced by its density.
Therefore, the above results suggest that the mechanical properties of the composites may
be dependent on the processing conditions since density differences exist in the face and
core region of the panels due to the manufacturing process. Since the overall density was
relatively the same between the panels with extruded and unextruded wood particles,
differences in mechanical properties associated with the processing conditions might be

attributed to the density profile in the panels.

Effects of Processing Conditions

The modulus of rupture (MOR), modulus of elasticity (MOE), internal bond (IB)
strength and screw holding capacity data for panels manufactured in this study, along
with standard requirements for particleboard of medium density grades, ranging from
640-800 kg/m3, are presented in Table 5.5. To determine significant differences between
processing conditions, comparisons were made by varying the processing conditions
(rows 1 vs. 2 and 3 vs. 4 under experimental panels) while holding the maleated

polyolefin type constant.

137

Table 5.5.

Effects of processing methods and material compositions on the
mechanical properties of particleboard panels bound with maleated
polyolefins.

 

Mechanical Properties1

 

 

 

 

 

Panel
Types MOR MOE IB Strength Screw
(MPa) (MPa) (MPa) Holding (N)
Medium
Density 11.0 — 16.5 1725 — 2750 0.40 - 0.55 900 — 1100
Grades1
Experimental Panels2
MAPE— 25.41 :30A 2068i233A 1.22:0.32A 1353 i184A
unextruded
MAPE“ 20.70i3.4B 1296:195B 2.07:0.69B 1563: 180A
extruded
MAPP - 30.04 i 6.6 A 3582 i 567C 0.43 : 0.19C 1469 i 330 A
unextruded
MAPP- 23.00:4.7B 2875 i347D 1.50:0.29D 1580:299A
extruded

 

lProperty requirement data is from standard ANSl A208.1- 1999-Particleboard.

2The means with different letters indicate significance between treatments at the a = 0.05
level, while the means with the same letter indicates no difference between treatments.

138

Mechanical properties of the composite panels differed depending on
manufacturing processes. Panels had significantly higher MOR and MOE values when
the wood particles were not extruded prior to pressing. This was likely due to localized
melting of the unreacted maleated compounds and greater flow at the faces of the panels,
causing compaction in the face region, due to the direct heat from the platens. The faces
would see more heat throughout the pressing cycle, likely causing the reaction between
the wood and binding agent as well. Greater compaction of the face region of the panels
was supported by the higher density of that region, as determined through X-ray density
profile analysis (Table 5.4). High face region density has been correlated to increased
MOR and MOE in conventional particleboard made with formaldehyde-based adhesives
(17, 18). In addition, increasing density is related to increased strength properties
through a power law relationship (19). Since overall density was relatively the same
between the panels with extruded and unextruded wood particles, the increased face
density of panels with unextruded wood particles must be responsible for the enhanced
bending properties of these panels. Although panels made with extruded wood particles
had lower bending properties than their unextruded counterparts, these panels still

exceeded the requirements for conventional particleboard in most cases.

Table 5.5 also summarizes the IB strength of composite panels under different
manufacturing methods. Internal bond strength is an indication of how well the particles
are bonded in the panel, particularly at the core region. The experimental results
indicated that the unextruded wood particles underwent both the grafting reaction and
entanglement during compression molding since panels were successfully produced

without pre-reacting wood particles in the extruder. Unlike the bending properties,

139

 

panels prepared from unextruded wood particles had lower IB strength, compared to
those made from extruded wood particles. Therefore, the lesser lB strength of composite
panels with untreated wood particles was attributed to the reduced density in the core
region of these boards, due to the heat not flowing to the center of the panel fast enough
to cause adequate polymer flow and panel compaction during the limited pressing time.
Similar trends have been reported for conventional glued particleboard (17, 18). In
addition, heat is required to drive the reaction between the wood particles and the
maleated polyolefins. Since the unextruded wood particles experience heat only during
the hot pressing step, they may not receive sufficient heat to complete the esterification
reaction and form chemical bonds between the wood and maleated polyolefins, especially
in the core region of the panels. By contrast, wood particles that were pre-reacted in the
extruder likely had more extensive bonding due to the extra heat and mixing during the
extrusion step. This accounts for the significantly higher IB strength in panels made with

extruded wood particles.

Processing methods had no effect on the screw holding capacity, which was

higher than the requirements for particleboard of medium density (Table 5.5).

Effects of Binding Agent Compositions

Two comparisons were made to determine significant effects of maleated
polyolefin base resin types (PE vs. PP) (Table 5.5), and maleic anhydride content in
maleated polypropylene (1.5% vs. 2.5% by weight)/molecular weight (52,000 vs. 47,000

g/mol) on the mechanical properties of the composites (Table 5.6).

140

 

Table 5.6.

Effect of molecular weight/maleic anhydride content of MAPP on the
mechanical properties of particleboard panels bound with maleated

polypropylenes

 

Mechanical Properties1

 

 

 

 

 

Panel
Types MOR MOE 1B Strength Screw
(MPa) (MPa) (MPa) Holding (N)
Medium
Density 11.0 — 16.5 1725 - 2750 0.40 - 0.55 900 — 1100
Gradesl
Experimental Panels2
MAPP
03003 30.04 i 6.6 A 3582 i 567 A 0.43 2: 0.19 A 1469 i 330 A
unextruded
MAPP
03015 30.24 i 8.6 A 3586 i 698 A 0.36 i 0.16 A 1445 i157 A
unextruded
MAPP A
3-3015 1986:49B 2663 :270B 1.60 :0.64B 1552 i193
extruded
MAPP B A
@3003 23.0 0: 4.7 B 2875 i 347 B 1.50 _+_ 0.29 1580 i 299
extruded

 

lProperty requirement data is from standard ANSI A208.1- l999-Particleboard.

2The means with different letters indicate significance between treatments at the or = 0.05
level, while the means with the same letter indicates no difference between treatments.

141

 

For the comparison of PE vs. PP-based binding agents, the specific comparisons
made in Table 5.5 were rows 1 vs. 3 and rows 2 vs. 4. In this way, processing condition
was held constant, while the polymer base resin was varied. Composite panels with
polypropylene-based binding agents outperformed their polyethylene counterparts in
stiffness (MOE), regardless of manufacturing method used, mainly due to the higher
stiffness of polypropylene in the binding agent (Table 5.5). However, the strength of the
composites (MOR) was not affected by the type of base resin of the maleated polyolefin
since both PE and PP behave similarly, regardless of the processing method. Conversely,
panels made with maleated polyethylene outperformed those made with maleated
polypropylene in IB strength, likely due to the lower melting temperature of
polyethylene. Lower melting temperature would allow » the polyethylene-based
compound to flow to a greater extent even into the center region of the panels, causing
stronger internal bonding. The screw holding capacity of the panels was not affected by

the base resin type in maleated polyolefin.

Differences between the two maleated polypropylene compounds were not
significant for any of the mechanical properties tested (Table 5.6). Specific comparisons
made were row 1 vs. 2 and row 3 vs. 4, where processing condition was held constant
while the molecular weight/maleic anhydride content of MAPP was varied. The weight
average molecular weights of the two MAPPs differed only by 5,000 g/mol and the
difference in maleic anhydride content was 1% between the two. These polymers may
have been similar enough that they did not create significant differences in the composite

panel properties.

142

Comparison with Standard ANSI A208.]

The MOR, IB strength and screw holding capacity results for experimental panels
manufactured in this study indicated that the standard requirements were met or
surpassed for all grades of particleboard of medium density when the particles were
extruded before pressing (Tables 5.5-5.6). Without the extrusion step, the 18 strength
was within the required range with maleated polypropylene, and surpassed when PE-
based binding agent was used. MOE data were below the standard requirements for
stiffness when MAPE was used with extrusion, but the panels with unextruded wood
particles bonded with MAPE surpassed the stiffness requirements. Additionally, when
MAPP was used, the panels exceeded the requirements for all grades of particleboard of

medium density, regardless of processing conditions.

143

 

CONCLUSIONS

This study examined the differences between panels made with MAPE and
MAPP binding agents, specifically the contrasts of: (i) base resin type, PE or PP, and (ii)
molecular weight/maleic anhydride content in MAPP binding agents, along with the
effects of the manufacturing process. Particles were either pre-reacted with maleated
binding agents in the extruder and then compression molded in a two-step process, or

directly compression molded without the extrusion step.

Surface characterization of unmodified and modified wood particles after reactive
extrusion with maleated polyolefins showed similar changes in the wood particle surfaces
after reaction, regardless of maleated polyolefin used. This evidence suggested that the
wood particles can undergo the same reaction in a one-step process, without any obvious

differences in reactivity.

Differences in mechanical properties of the panels were correlated to the type of
base resin used in the binding agent and to panel density profile. Results showed that
while extruding the particles before panel pressing gave better overall internal bond
strength, superior bending properties were obtained through compression molding alone.
MAPP based panels outperformed MAPE based panels in stiffness, likely due to the
higher stiffness of the PP base resin. MAPE enhanced the IB strength compared to
MAPP, attributed to better melting and flow of the polyethylene. Polymer base resin had
no effect on MOR or screw holding capacity. Differences between the two maleated
polypropylene compounds were not significant for any of the mechanical properties

tested.

144

The study also showed that a new type of environmentally friendly wood
composite product could be formed from modified wood particles, regardless of
processing conditions. This composite contained no formaldehyde-based adhesive, but
still performed very well in mechanical tests, in many cases exceeding the standard

requirements for particleboard of medium density.

145

 

10.

11.

REFERENCES

Matsuda, H., "Preparation and utilization of esterified woods bearing carboxyl
groups," Wood Science and Technology, 21 (1): 75-88 (1987).

Hon, D.N.S. and Xing, L.M., "Thermoplasticization of wood. Esterification," in
Viscoelasticity of Biomaterials, W.G. Glasser and H. Hatakeyama, Editors.
American Chemical Society: Washington DC. p. 118-132 (1992).

Clemons, C., Young, R.A., and Rowell, R.M., "Moisture sorption properties of
composite boards from esterified aspen fiber," Wood and Fiber Science, 24 (3):
353-63 (1992).

Marcovich, N.E., Reboredo, M.M., and Arangguren, M.I., "Sawdust modification.
Maleic anhydride chemical treatment," Holz als Roh- und Werkstojf, 54 (3): 189-
193 (1996).

Marcovich, N.E., Aranguren, M.I., and Reboredo, M.M., "Modified woodflour as
thermoset fillers Part I. Effect of the chemical modification and percentage of
filler on the mechanical properties," Polymer, 42 (2): 815-825 (2001).

Marcovich, N.E., Reboredo, M.M., and Aranguren, M.I., "Modified wood flour as
thermoset fillers. 11. Thermal degradation of wood flours and composites,"
T hermochimica Acta, 372 (1-2): 45-57 (2001).

Matsuda, H., Ueda, M., and Mori, H., "Preparation and crosslinking of
oligoesterified woods based on maleic anhydride and allyl glycidyl ether," Wood
Science and Technolog, 22 (1): 21-32 (1988).

Matsuda, H., "Thermal plasticization of lignocellulosics for composites," in
Emerging Technologies for Materials and Chemicals from Biomass R.M. Rowell,
T.P. Schultz, and R. Narayan, Editors. American Chemical Society: Washington,
DC. p. 98-114 (1992).

Timar, M.C., Mihai, M.D., Maher, K., and Irle, M., "Preparation of wood with
thermoplastic properties. Part 1. Classical synthesis," Holzforschung, 54(1): 71-
76(2000)

Timar, M.C., Maher, K., Irle, M., and Mihai, M.D., "Preparation of wood with
thermoplastic properties. Part 2. Simplified technologies," Holzforschung, 54 (1):
77-82 (2000).

Timar, M.C., Maher, K., Irle, M., and Mihai, M.D., "Thermal forming of
chemically modified wood to make high-performance plastic-like wood
composites," Holzforschung, 58 (5): 519-528 (2004).

146

 

 

 

12.

13.

14.

15.

16.

17.

18.

19.

Carlbom, K. and Matuana, L.M., "Composite materials manufactured from wood
particles modified through a reactive extrusion process," Polymer Composites, 26
(4): 534-541 (2005).

Carlbom, K. and Matuana, L.M., "Functionalization of Wood Particles through a
Reactive Extrusion Process," Journal of Applied Polymer Science: In press
(2006)

ASTM, D 1037-99, Standard Methods of Evaluating the Properties of Wood-
Based Fiber and Particle Panel Materials. ASTM: West Conshohocken (1999).

ANSI, A208.1-1999, Particleboard. The Composite Panel Association:
Gaithersburg (1999).

Ratner, 8D. and Castner, D.G., "Electron Spectroscopy for Chemical Analysis,"

in Surface Analysis: The Principal Techniques, J .C . Vickerman, Editor. John
Wiley & Sons: Chichester, England. p. 43-98 (1997).

Schulte, M. and F ruhwald, A., "Some investigations concerning density profile
internal bond and relating failure position of particle board," Wood Science and
Technology, 54: 289-294 (1996).

Wong, E.D., Zhang, M., Wang, Q., and Kawai, 8., "Formation of the density
profile and its effects on the properties of particleboard," Wood Science and
Technology, 33 (4): 327-340 (1999).

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

147

 

 

 

CHAPTER 6

MODELING AND OPTIMIZATION OF F ORMALDEHYDE-FREE WOOD

COMPOSITES USING A BOX-BEHNKEN DESIGN

This chapter has been accepted for publication in Polymer Composites (February 2006).
It is co-authored by K. Carlbom and L. M. Matuana

148

 

ABSTRACT

A response surface model using a Box-Behnken design was constructed to
statistically model and optimize the material compositions-processing conditions-
mechanical property relationships of formaldehyde-free wood composite panels. Three
levels of binding agent content, pressing time and press temperature were studied and
regression models were developed to describe and optimize the statistical effects of the
formulation and processing conditions on the mechanical properties of the panels. Linear
models best fit both the flexural strength (MOR) and internal bond (IB) strength of the
panels. Increasing any of the manufacturing variables resulted in greater MOR and IB
strength. F lexural stiffness (MOE) was best described by a quadratic regression model.
Increased MOE could be obtained through higher pressing times, binding agent
concentrations and/or pressing temperatures. However, binding agent concentration had
less effect on increasing the MOE at higher pressing temperatures. Numerical
optimization showed that formaldehyde-free panels with desired mechanical properties
could be manufactured at pressing temperatures ranging from 187.18—199.97°C, pressing

time from 3.31—8.83 minutes, and binding agent concentration from 7.66—11.86%.

149

 

 

 

INTRODUCTION

The recent implementation of more stringent emissions regulations for
manufacturing facilities that produce particleboard and other wood composite panels (1)
has spurred research into alternative binding technologies that do not involve
formaldehyde. Our prior work demonstrated the efficiency of the reactive extrusion
process as a way to graft maleated polyolefins onto the surface of wood particles, which
could then be bound together without any additional adhesive (2-4). A new type of
environmentally-friendly wood composite product was manufactured from the modified
wood particles. This composite contained no formaldehyde-based adhesive, but still
performed very well in mechanical tests, in some cases exceeding the standard
requirements for particleboard of medium density.

In previous work, the effects of extrusion processing conditions on the grafting of
maleated polyolefins to wood particles were investigated, but work is still needed to
determine the relationships between the mechanical properties of the formaldehyde-free
composite panels and the material composition and panel manufacturing variables (2-4).

The quality of wood composite materials such as particleboard depends heavily
on material composition and manufacturing condition variables such as resin content,
pressing temperature and pressing time (5). When multiple variables are involved, it
becomes difficult to study the system using the common approach of varying only one
factor at a time, while holding the others constant. This approach is not only time-
consuming, but can also be costly and does not easily identify all of the interactions
between factors (6, 7). A more efficient way to investigate these systems is to develop a

mathematical model describing the relationship between the response and independent

150

 

variables, in which the significance of individual factors and multi-factor interactions can
be determined (6-9).

A Box-Behnken design (BBD) is a versatile method to statistically model and
optimize response variables that are affected by multiple independent factors. Compared
to full factorial or central composite designs, the BBD requires fewer trials and can
efficiently model quadratic or higher-order relationships. Because of these features, the
BBD has been used to study a variety of wood composite products (10-13).

In this study, the effects of binding agent content, pressing time and pressing
temperature on the mechanical properties of the resulting panels were evaluated using a
Box-Behnken design to statistically model the system. Mechanical properties studied
included modulus of rupture (MOR), modulus of elasticity (MOE) and internal bond (IB)
strength, all of which are key parameters for assessing panel quality. Numerical
optimization was performed in order to determine the best conditions for the manufacture

of formaldehyde-free wood composite panels.

151

 

EXPERIMENTAL
Materials

Maple wood particles of 425 micron (40-mesh) size used in this study were
supplied by American Wood Fibers (Schofield, WI). Hydrated zinc acetate, the catalyst,
was obtained from Baker Analytical Reagents (JT Baker Co., Phillipsburg, NJ).
Maleated polypropylene (MAPP or G-3003) supplied by Eastman Chemical Co.
(Kingsport, TN) was used as the binding agent. The MAPP had a weight average
molecular weight (MW) of 52,000 g/mole, approximate viscosity of 60,000 cP at 190°C

and maleic anhydride content of 1.5% by weight.

Experimental Design

The Box-Behnken design (BBD) is a three-level design based upon the
combination of two-level factorial designs and incomplete block designs (6, 7). BBDs
are spherical designs, with the design points for high and low levels located at an equal
distance from the center of the design. These designs have excellent predictability within
the spherical design space and require fewer experiments than full factorial designs or
central composite designs (CCDs) with the same number of factors. For example, an
investigation with three factors would require at least 27 experiments in a full factorial
design, 15 experiments in a CCD, or 13 experiments in a BBD, with additional replicates
of the center point as necessary in each design to estimate experimental error.
Additionally, BBDs are rotatable or nearly rotatable regardless of the number of factors

studied (6, 7).

152

All statistical analysis, modeling and numerical optimization was performed using
Design Expert software, v.6 (Stat-Ease, Inc. Minneapolis, MN). The BBD matrix
generated by Design Expert software displays factor levels in the experimental design in
two ways (i) the actual factor levels, which are the values from the experiment, and (ii)
the coded factor levels, +1, -1, and 0, for high levels, low levels, and center point,

respectively. Coded factor levels are defined as:

Coded factor levels = Actual value - Factor mean (1)
(Range of factorlal values/2)

 

The BBD experimental design matrix is shown in terms of both actual and coded

factor levels in Table 6.1. Twelve replicates were run for each experiment.

153

Table 6.1. Box-Behnken Design Matrix in terms of Both Actual and Coded Factor
Levels Generated by Design Expert Software.

 

 

 

Factors
Effifgaim goint A: Press B: Press Time C:CBinding Agent
ype Temperature (°C) (min) onchrtlon
1 IBFact 160 (-1) 6 (0) 18 (+1)
2 Center 180 (0) 6 (0) 10.5 (0)
3 IBFact 180 (0) 3 (-l) 18 (+1)
4 IBFact 160 (-1) 9 (+1) 10.5 (0)
5 IBFact 160 (-1) 6 (0) 3 (-1)
6 IBFact 180 (0) 9 (+1) 18 (+1)
7 IBFact 180 (0) 3 (-l) 3 (-1)
8 IBFact 200 (+1) 9 (+1) 10.5 (0)
9 IBFact 200 (+1) 6 (0) 3 (-1)
10 IBFact 200 (+1) 6 (0) 18 (+1)
11 IBFact 160 (-l) 3 (-1) 10.5 (0)
l2 IBFact 180 (0) 9 (+1) 3 (-1)
13 IBFact 200 (+1) 3 (-1) 10.5 (0)

 

154

 

Compounding and Panel Manufacture

The wood particles were dried for 48 hours at 105°C to a final moisture content of
less than one percent before processing. A 10-liter high intensity mixer (Papenmeier
TGAHKZO) was used for dry blending of the wood particles, binding agent, and catalyst.
Levels of binding agent were varied as indicated in Table 6.1, based on the oven dry
weight of wood flour. Zinc acetate esterification catalyst was added at 1% of the binding
agent weight in all cases. All components were combined in the mixer and blended for
10 minutes at room temperature. Following blending, reactive extrusion was used to

induce the esterification reaction between maleated polypropylene and wood particles.

This was achieved as follows: the compounded wood particles were fed into a 32
mm conical counter rotating twin-screw extruder (C. W. Brabender Instruments, Inc.)
with an L/D ratio of 13:1, driven by a 7.5 hp Intelli-Torque Plasti-Corder Torque
Rheometer®. The barrel temperatures for all three zones inside the extruder were set at

165°C, and the rotational speed of the screws was held at 60 rpm.

Once extruded, the wood particles were compression molded into panels using 3.4
MPa of pressure at various pressing times and temperatures set at levels indicated in
Table 6.1. After pressing, panels were removed from the press and cooled at room
temperature under compression for 15 minutes. Panel dimensions were 380 by 380 by 6

mm, with a target density of 750 kg/m3.

155

 

Property Testing

Three-point flexural and internal bond (IB) strength tests were performed on an
Instron 4206 testing machine (using Series IX software) in accordance with procedures
outlined in ASTM standard D1037-99 (14). The crosshead speeds were 3.05 mm/min and
8.13 mm/min for flexural and IB tests, respectively. Values for modulus of rupture
(MOR), modulus of elasticity (MOE) and IB strength were compared with values listed

for particleboard of medium density in the standard ANSI A208.1-1999 (15).

156

 

RESULTS AND DISCUSSION

Mechanical Properties

To determine whether experimental panels would conform to standard strength
requirements for conventional particleboard, mechanical property data was compared to
ANSI standard requirements for particleboard. Table 6.2 lists the MOR, MOE, and IB
strength requirements for particleboard of medium density, ranging from 640-800 kg/m3.
In previous studies, composite panels manufactured through this process were within this
range, with average densities ranging from 775-782 kg/m3 (2, 4). There are four grades
of particleboard of medium density, all of which can be made with either interior or
exterior adhesives. Grades M-1 and M-S are commercial grade boards, while M-2 and

M-3 are intended for industrial use.

Results of MOR, MOE and IB strength tests revealed a large range of property
values for the experimental panels (Table 6.2). At the lowest levels of pressing time,
temperature and binding agent concentration, panels were below the standard property
requirements. However, at the highest levels of these factors, most panels met the
standard requirements for particleboard of medium density. Several panels even
exceeded the standard requirements for various grades of particleboard. These results are
significant because particleboard is currently manufactured with formaldehyde-based
adhesives. The formaldehyde-free wood composites manufactured in this study are more
environmentally friendly and often outperform the requirements listed in the standard

ANSI A208.].

157

Table 6.2. Standard Property Requirements for Various Grades of Particleboard of
Medium Density (640-800 kg/m3)

 

 

 

 

Gradesl Expleglrlzgtal
Properties
M-l M-S M-2 M-3 Property Range
MOR (MPa) 11.0 12.5 14.5 16.5 0.48 — 17.91
MOE (MPa) 1725 1900 2250 2750 101.2 — 2556
IB (MPa) 0.40 0.40 0.45 0.55 0.04 — 2.69

 

‘From Standard ANSI A208.1-1999 Particleboard.

2Experimental values are from panels manufactured under the various factor
combinations specified in Table 1. Minimum values are from panels pressed
at 160°C for 6 minutes with 3% MAPP, while maximum values are from
panels pressed at 200°C for 6 or 9 minutes with 10.5 or 18% MAPP.

158

 

Statistical Analysis of the Model

Regression analysis was performed on the mechanical property results in order to
develop best-fit models for the experimental data. A separate regression analysis was run
for each of the three mechanical properties studied to determine the relationships between

mechanical properties and binding agent content, pressing time and pressing temperature.

Modulus of Rupture (MOR) and Internal Bond (IB) Strength

For both MOR and IB strength, regression analysis of the experimental data
showed that a linear model best fit the relationship between the response (MOR or IE)
and the binding agent content, pressing time and pressing temperature. In both cases, a
square root transformation was applied to the data in order to normalize the variance of

the residuals. These models, described in terms of coded factors, were described as:

4MOR = +2.75 + 0.78A + 0.23B + 0.91C (2)
and
4/IB strength = +1 + 0.30A + 0.084B + 0.54C (3)

where A is press temperature, B is press time, and C is binding agent concentration.

For both MOR and IB strength, only the three main factors were significant, with
no interactions found between them. Since the equations are displayed in terms of coded
factors, the relative effect of each variable on the response can be evaluated by

comparing the absolute value of its coefficient and its algebraic sign. As shown in

159

 

Equations 2 and 3, the effects of all of the factors had positive algebraic signs, which
indicated that increasing any one would increase theMOR or IB strength. In both cases,
concentration of binding agent (factor C) had the largest effect on the mechanical
properties, followed by pressing temperature (factor A) and pressing time (factor B).
This is illustrated by a cube graph in Figure 6.1(a) for MOR and 6.1(b) for IB. Increasing
the amount of binding agent resulted in better adhesion between the wood particles as a
result of having more polymer chains attached to the wood particles which would
entangle to hold the panel together. Similarly, an increase in temperature would allow
more thorough melting and flow of the polymer chains attached to the wood particles,
thus increasing the efficiency of binding. Although longer pressing time did increase
both MOR and IB, the effect was much smaller which indicated that pressing time was
much less important in improving the MOR and IB than binding agent concentration or

press temperature.

160

 

(a)
Sqrt( MOR)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3.11 4.68
B+ 1.29 2.85
G)
E
l:-
g 2.64 4.21 C+
o.
in
C: Concentration
B- 0.82 2.38 C-
A- A+
A'Temperature
(b)
Sqrt(lB)
1.32 1.92
B+ 0.24 0.85
(D
E
1:
8
9 1.15 1.75 C+
a.
63'
C: Concentration
B- 0.08 0.68 C-
A- A+

A‘ Temperature

Figure 6.1. Cube graphs of the linear relationship between mechanical property results
and press temperature, pressing time, and binding agent concentration for

(3) MOR and (b) IB strength.

161

Modulus of Elasticity (MOE)

Results of the regression analysis for MOE suggested that the data was best fit by
a quadratic model. A. square root transformation was also applied to the data for MOE in
order to normalize the variance of the residuals in the system. Table 6.3 shows the
analysis of variance (ANOVA) results for the response surface quadratic model. As
listed in the table, some of the factors were not significant (Probability > F greater than
0.0500), so the model was reduced by removing the insignificant factors. The reduced

model, containing only significant terms and described in terms of coded factors is:

JMOE = + 33.49 + 9.47A + 3.153 + 9.99C +1.14A2 + 2.7032 - 2.44C2 - 2.02AC (4)

For MOE, all main effects (A, B and C) were significant, along with the second-
order main effects (A2, B2 and C2), and the AC interaction. Because A, B and C had
positive algebraic signs, an increase in temperature, pressing time or binding agent
concentration would improve the MOE. By contrast, the two-factor AC interaction had a

negative algebraic sign, which indicated a negative effect on MOE.

162

 

Table 6.3. Analysis of Variance (ANOVA) for Response Surface Quadratic Model

 

 

Source 88:31:; lifrgefifslzf S1232: F-Value Prob. > Fl

Quadratic Model 1671.35 9 185.71 379.34 < 0.0001

A 716.70 1 716.70 1463.99 < 0.0001

B 79.35 1 79.35 162.08 < 0.0001

C 799.04 1 799.04 1632.19 < 0.0001
A2 5.49 1 5.49 11.21 0.0123

B2 30.59 1 30.59 62.49 < 0.0001
C2 25.00 1 25.00 51.07 0.0002
AB 0.58 1 0.58 1.18 0.3137
AC 16.30 1 16.30 33.30 0.0007
BC 0.89 1 0.89 1.83 0.2187

 

1Values of “Prob. > P” less than 0.0500 indicate model terms are significant. In this case
A, B, C, A2, B2, C2 and AC are the significant model terms.

163

 

The perturbation plot of the square root of MOE against temperature, pressing
time and binding agent concentration shows the contribution of each factor to the MOE
(Figure 6.2). The perturbation plot illustrates the changes in MOE as each factor moves
from the chosen reference with all other factors held constant at the middle level of the
design space (6, 7). Pressing temperature and binding agent content (factors A and C) are
shown to have the largest effect on the MOE, while pressing time (factor B) shows only a

small effect on MOE.

Because two of the main factors (A and C) that increase the MOE are part of a
significant interaction, it would not be appropriate to investigate these factors separately.
The effect of one factor will depend upon the level of the other, since the interaction is

significant.

164

 

 

Figure 6.2.

Sqrt(MOE)

 

 

 

 

 

51—
A
40—
L
30—
20—
10
l l l l l
-1.0 -05 0.0 0.5 1.0

Deviation from Reference Point

Perturbation plot of square root of MOE against pressing temperature (A),
pressing time (B) and binding agent concentration (C).

165

 

Figure 6.3 shows the interaction graphs of the change in MOE as a function of the
temperature-concentration interaction. The effect of the temperature-concentration
interaction is illustrated at the lowest pressing time, 3 minutes, in Figure 6.3(a) and at the
highest pressing time, 9 minutes, in 6.3(b). A small increase in MOE resulted from
increased pressing time, regardless of the temperature or binding agent concentration.
However, this effect was quite small, resulting in gains of only 3-8 MPa to the MOE over

the range studied.

The binding agent concentration had greater effect on MOE when the pressing
temperature was the lowest. At higher temperatures, the differences between low and
high binding agent concentration were reduced. This can be verified by comparing the
magnitude of the difference between the lines for high and low binding agent
concentration. At the lowest temperature, the difference in MOE was large, but was
reduced as the temperature increased. This reduced efficiency of binding agent
concentration to increase the MOE at higher temperatures accounts for the negative effect
of the temperature-binding agent interaction, which was shown by the negative algebraic
sign on AC in Equation 4. The effect of this interaction is that less binding agent is
required at higher temperatures to produce panels with sufficient MOE values. At lower
temperatures, more binding agent would be required to manufacture composites with the
same range of MOE values. A greater degree of melting and flow of the polymer
attached to the wood particles likely occurs at higher temperatures, which would allow a
smaller amount of binding agent to be as more efficient in creating particle-particle

bonds.

166

 

Figure 6.3.

 

 

 

 

 

 

 

 

60- (a)
18%
47- \
ft?
3 34
E’
O'
a)
21-1
3%
3..
l I T l l
160 170 180 190 200
Temperature °C
60- (b)
18%
47-
1’1?
2 3
4-1
E"
U
m
a-
l l I I I
160 170 180 190 200

Temperature °C

Interaction plots of the variation in square root of MOE as a function of
the interaction between pressing temperature and binding agent
concentration at (a) low press time (3 minutes) and (b) high press time (9
minutes).

167

 

Numerical Optimization of Mechanical Properties

The numerical optimization function in Design Expert software was used to
determine combinations of binding agent concentration, panel pressing time and
temperature that would result in the most favorable mechanical properties of the resulting
composites. The numerical optimization function of Design Expert is based on the
desirability function developed by Derringer and Suich (16), which transforms each
response value to a desirability index (di). Each desirability index is defined by three
parameters: goal, lower and upper, and the program allows the d, goal parameter to be to
one of five options: minimum, maximum, target, in range, or equal to. Once these
settings have been defined, d; varies between zero (worst case) and one (ideal case).
Design Expert searches for the largest overall di and presents a series of solutions which

best maximize the di.

Table 6.4 lists the optimization criteria settings used to optimize the mechanical
properties of the composite panels. In order to produce a composite panel that had
adequate strength (MOR), stiffness (MOE), and IB strength, the goal for each property
was initially set in range based upon the property requirements of the particleboard
standard ANSI A208.] (Table 6.2). For example, square root of MOR was set in range
3.32—4.06 MPa, MOE was set in range 41 .53—52.44 MPa and IB strength was set in range
0.63—0.74 MPa (15). However, the data for IB strength showed that the standard
property requirements were met and even exceeded at very low concentrations of binding
agent. Since the IB would be higher than required at nearly all combinations of pressing
temperature, panel pressing time and binding agent concentration, the IB criteria was set

in range over all experimental values instead of in range of the property requirement

168

 

values. In this way, the MOR and MOE could be targeted to the necessary values, with
the understanding that IE strength would likely be much higher than required.
Additionally, because we wanted to determine which combination of material
composition and processing conditions would produce panels with the desired properties,
the optimization criteria for all of these factors were set in range over the entire set of
studied values. For this analysis, the desirability functions of the mechanical properties

were set as follows:

(i) if JMOR < 3.32 or JMOR > 4.06 MPa, and x/MOE < 41.53 or
J MOE > 52.44 MPa then (I, = 0 (worst case)

(ii) if 3.32 S «lMOR S 4.06 and 41.53 S \lMOE S 52.44, then (1, = 1 (ideal
case)

All other optimization parameters were set to the default settings of one for the
weights of upper and lower limits for the input factors, and three for the importance (3
relative scale that weights each of the resulting dis in the overall desirability product).
Ten cycles were run per optimization, with the epsilon value for the minimum difference

in eliminating duplicate results set at its default value.

169

 

 

 

 

 

 

 

Ff. l .11 iii.
oo: o: 2.: Sam NEW cfiw :43— o
co: em: ”5:4 nmd mam: m3. Neda— w
co: em: 3.? owm 3.: no.» _méo— n
co: mm: cart? ova 3.: 3.x 2.3— o
co: m _ ._ code mm.m mod 36 mada— m
co: om: mode 3am mm: _ Sam 9462 :0
co: mm: 3.? om.m 2.: _m.m 3.93 m
co: N. _ ._ no.3 N.Mm obs 3.x 360— N
oo: o: 3.34 mm.m cos 2.» 3.3. _
3:6 :55 3:»: 3L 5:838:60 :55 Ge
E328: Eat—em Eozvtcm 226:5 so? “seam 65.9 8o...— 23885oh 8:28
mucus—em 83330

m _ _ 8.. one owes 5 Es: 83:5

m _ _ 3.2 mm. _ a owes 5 3:6 Amozeem

m _ _ 2:. mg once 5 And—2:102:23

m _ _ w: m awash E ax; 5:828:00 Eow< wEvEm

m _ _ o m emcee E AEEV oEC. @535

m _ _ com of 098.. E Gov 23anth wfimmoi

sue? an?»
agar—3:: .59:— 333 :85 ~23: :85 33:4— 130 .885—

 

mafiahmacv 8:27.3—

 

mzamom can mwfifiom 5:23.530 _motofizz

66 03mm.

170

Numerical optimization results produced 9 optimum solutions, all with
desirability of 1.00. These results are displayed in Table 6.4. As listed in the table,
pressing temperatures ranged from 187.18—199.97°C, pressing time ranged from 3.31—
8.83 minutes, and binding agent concentration from 7.66—11.86%. These results showed
that formaldehyde-free wood composite products could be made under a variety of
conditions, many of which are not all that different than those used to manufacture
conventional particleboard. In addition, a desired solution could be employed if the goal
was to minimize a particular factor, such as pressing time, in order to reduce

manufacturing costs.

171

 

CONCLUSIONS

Three panel manufacturing variables (pressing temperature, pressing time and

binding agent content) were analyzed by developing statistical models for their

relationships with mechanical properties MOR, MOE and IB strength. Additionally,

numerical optimization was used to determine the best conditions for manufacturing the

panels in terms of the mechanical properties. The following conclusions were obtained:

1.

Both MOR and IB were best fit with linear models. Increasing any of the
manufacturing variables (pressing temperature, pressing time and binding agent

content) resulted in greater MOR and IB strength.

A quadratic model best fit the MOE. Increased MOE could be obtained through
longer pressing times, but the effect of this variable was small. Larger effects on
MOE were obtained by increasing the binding agent concentrations and pressing
temperatures. However, the negative interaction between temperature and
binding agent concentration indicated that binding agent concentration had less

effect on increasing the MOE as the pressing temperature increased.

Panels with mechanical property values in the range of the mechanical property
requirements for particleboard of medium density could be produced using
several pressing temperature-pressing time-binding agent concentration
combinations with maximum desirability. This showed that formaldehyde-free
panels could be produced at similar adhesive contents and processing conditions
used for conventional glued particleboard made with formaldehyde-based

adhesive.

172

 

 

 

10.

11.

12.

REFERENCES

US-EPA, "National emission standards for hazardous air pollutants: plywood and
composite wood products," Federal Register, 69 (146): 45943-46046 (2004).

Carlbom, K. and Matuana, L.M., "Composite materials manufactured from wood

particles modified through a reactive extrusion process," Polymer Composites, 26
(4): 534-541 (2005).

Carlbom, K. and Matuana, L.M., "Functionalization of Wood Particles through a
Reactive Extrusion Process," Journal of Applied Polymer Science: In press

(2006)

Carlbom, K. and Matuana, L.M., "Influence of Processing Conditions and
Material Compositions on the Performance of Formaldehyde-Free Wood-Based
Composites," Polymer Composites: In press (2006).

Maloney, T.M., Modern Particleboard and Dry-Process F iberboard
Manufacturing, Updated Ed., Miller Freeman: San Francisco (1993).

Myers, RH. and Montgomery, D.C., Response Surface Methodology: Process
and Product Optimization Using Designed Experiments. John Wiley & Sons, Inc.:
New York (1995).

Montgomery, D.C., Design and Analysis of Experiments, 5th Ed., John Wiley &
Sons, Inc.: New York (2001).

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

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

Barry, A., Lepine, R., Lovell, R., and Raymond, 8., "Response surface
methodology study of VOCs in plywood press emissions," Forest Products
Journal, 51 (1): 65-73 (2001).

Barry, AD. and Comeau, D., "Volatile organic chemicals emissions from OSB as
a function of processing parameters," Holzforschung, 53 (4): 441-446 (1999).

Barry, A., Comeau, D., and Lovell, R., "Press volatile organic compound
emissions as a function of wood particleboard processing parameters," Forest
Products Journal, 50 (10): 35-42 (2000).

173

 

13.

14.

15.

16.

Park, B.D., Riedl, B., Hsu, E.W., and Shields, J ., "Hot pressing process
optimization by response surface methodology," Forest Prod. J., 49 (5): 62-68
(1999)

ASTM, D 103 7-99, Standard Methods of Evaluating the Properties of Wood-
Based Fiber and Particle Panel Materials. ASTM: West Conshohocken (1999).

ANSI, A208.1-1999, Particleboard. The Composite Panel Association:
Gaithersburg (1999).

Derringer, G. and Suich, R., "Simultaneous optimization of several response
variables," Journal of Quality Technology, 12 (4): 214-219 (1980).

 

 

174

CHAPTER 7

SUMMARY OF FINDINGS

Public awareness about formaldehyde and other toxic chemicals being released to
the environment has helped to drive new government standards for wood composite
products. As environmental regulations become more stringent, the need for
formaldehyde-free adhesive systems for wood composite products will likely increase.
The main goal of this work was to study the concept of using a reactive extrusion process
as a means of developing a new, formaldehyde-free binding system for wood composite
products. The following specific objectives were accomplished to achieve the main goal

of this project:

1. Evaluate the effects of material composition (binding agent types and content)
and extrusion processing conditions (temperature profile and rotational screw
speed) on the level of grafting (surface properties) of wood particles alter the

reactive extrusion process;

2. Characterize the surface of unmodified and modified wood particles in terms of

chemical compositions (both elemental and functional groups);

3. Manufacture composites and evaluate their physico-mechanical properties;

175

 

4. Establish the relationships between material composition, processing conditions,
and physico-mechanical properties in order to identify the major factors that
govern the performance of formaldehyde-free wood composite products
manufactured through a reactive extrusion process.

The following sections relate the findings of this work to these objectives.

Objectives 1 and 2:

Results of FTIR, 13C NMR, XPS and titration analysis verified the reaction
between wood particles and maleated polyolefins. This proved that the maleated
polyolefins could be successfully grafted to wood particles using a reactive extrusion
process, without the use of any solvents. The esterification reaction between wood
particles and MAPE was found to be a function of the MAPE concentration used to
modify the wood particles. The grafting reaction produced mostly the diester form of the
modified wood particle during reactive extrusion. However, no significant difference
was found in grafting efficiency of the modified wood particles at different extrusion
processing conditions. Changing the extruder’s barrel temperature profile (BO-180°C)
and/or its rotational screw speed (20-80 rpm) resulted in adequate grafting, which

indicated that the esterification reaction was not a function of processing conditions.

Regardless of MAPP molecular weight (from a low of 11,200 to a high of 52,000
g/mol) all investigated MAPP compounds were effective in changing the surface of wood
particles after modification, compared to unmodified wood particles. However, no
distinct trend was observed between molecular weight of MAPP and grafting efficiency

through a reactive extrusion process. Regardless of molecular weight, it is believed that

176

 

the high content of MAPP (20%) used in this study prevented the detection of differences
in the grafting efficiency because the maximum level of grafting reaction had already

occurred with each component at 20% MAPP content.

The reactive extrusion process was found to be a suitable way to modify wood
particles with maleated polyolefins as it worked quickly and without the use of solvents.
This process would be industrially friendly and would allow large quantities of modified
wood particles to be produced for the manufacture of a formaldehyde-free wood

composite product.

Objective 3:

Modified wood particles made with MAPE were successfully formed into
formaldehyde-free composite panels without the use of any additional adhesive. The
composite contained no formaldehyde-based adhesive, but still performed very well in
mechanical tests, in many cases exceeding the standard requirements for particleboard of

medium density.

Additional work focused on the differences between panels made with MAPE and
MAPP binding agents, specifically the contrasts of: (i) base resin type, PE or PP, and (ii)
molecular weight/maleic anhydride content in MAPP binding agents, along with the
effects of the manufacturing process. Particles were either pre-reacted with maleated
binding agents in the extruder and then compression molded in a two-step process, or

directly compression molded without the extrusion step.

177

 

Differences in mechanical properties of the panels were correlated to the type of
base resin used in the binding agent and to panel density profile. Results showed that
while extruding the particles before panel pressing gave better overall internal bond
strength, superior bending properties were obtained through compression molding alone.
MAPP based panels outperformed MAPE based panels in stiffness, likely due to the
higher stiffness of the PP base resin. MAPE enhanced the IB strength compared to
MAPP, attributed to better melting and flow of the polyethylene. Polymer base resin had
no effect on MOR or screw holding capacity. Differences between the two maleated
polypropylene compounds were not significant for any of the mechanical properties

tested.

Objective 4:

In a final phase of the study, three panel manufacturing variables (pressing
temperature, pressing time and binding agent content) were analyzed by developing
statistical models for their relationships with mechanical properties MOR, MOE and IB
strength. Additionally, numerical optimization was used to determine the best conditions

for manufacturing the panels in terms of the mechanical properties.

Statistical modeling revealed that both MOR and IB were best fit with linear
models. Increasing any of the manufacturing variables (pressing temperature, pressing
time and binding agent content) resulted in greater MOR and IB strength. However, the
MOE was best fit by a quadratic model. Increased MOE could be obtained through
longer pressing times, but the effect of this variable was small. Larger effects on MOE

were obtained by increasing the binding agent concentrations and pressing temperatures.

178

 

However, the negative interaction between temperature and binding agent concentration
indicated that binding agent concentration had less effect on increasing the MOE as the

pressing temperature increased.

Panels with mechanical property values in the range of the mechanical property
requirements for particleboard of medium density could be produced using several
pressing temperature-pressing time-binding agent concentration combinations with
maximum desirability. This showed that formaldehyde-free panels could be produced at
similar adhesive contents and processing conditions used for conventional glued

particleboard made with formaldehyde-based adhesive.

179

 

Future Work

This work resulted in the development of a new type of formaldehyde-free wood
composite material that could be produced through a dry process. The surface
composition of unmodified and modified wood particles and maleated polyolefins were
studied in detail in order to verify that the reaction had occurred, and to determine the
efficiency of the modification under several different material compositions and
processing conditions. However, hardwood maple was used as the wood component in
nearly all of the work. Only one set of pine panels was manufactured in order to
determine a baseline comparison between wood species (see appendix). The effect of
wood species, specifically hardwood vs. softwood, should be studied in greater detail to
determine whether the wood type has an effect on the grafting efficiency during reactive
extrusion. The presence of naturally occurring resins in some softwood species might
have an effect on the grafting reaction, which may cause differences in the quality of
panels produced with this method. A large portion of particleboard is currently made
with softwood species such as pine, so this study would be relevant for industrial

application of the technology.

No differences were found in between modified wood particles produced under
various extrusion processing conditions when the concentration of MAPE was held
constant at 20%. The results of that portion of the study suggested that 20% by weight
might be too much binding agent to see any differences that may exist under these
conditions. In order to determine whether differences could be observed at lower binding
agent concentrations, a similar investigation should be run at a lower concentration,

perhaps between 5-10% by weight. This would be below the level that was determined to

180

 

cause maximum grafting, around 15%. If differences in grafting efficiency were found
with this lower concentration, the extrusion conditions could be modified for future work
in order to produce the greatest amount of grafting in the modified particles. Since this
work indicated that panels with properties within the standard requirement range could be
made at much lower concentrations than the 20% used in early studies, this would ensure

that the panel quality would be high when made with low binding agent concentrations.

In addition, the moisture resistance of wood composite panels such as
particleboard or MDF made with UF adhesive is generally very poor. Results of
thickness swelling tests after a 24 hour cold soak (see appendix) indicated that
formaldehyde-free composites had low swelling. Standard property requirement values
for particleboard and medium density fiberboard vary depending upon grade and density,
and not all grades specify thickness swelling requirements. Therefore, laboratory tests
should be performed using commercial particleboard or MDF as a control to determine
whether these formaldehyde-free panels have an advantage in moisture resistance. These
panels may prove to be better at resisting moisture than the conventional glued

particleboard and MDF, which would be another industrial selling point.

Finally, the studies done to determine whether panels with suitable mechanical
properties could be made through a one-step process (extrusion only) suggested that the
IB strength of these panels was lower than those made through the two-step process.
This was viewed as a drawback of this manufacturing method, even though the 1B
strength was close to or within the standard property requirement range from ANSI.
These panels were made at only one pressing condition, which had previously been

determined to produce panels with good mechanical properties in the two-step method.

181

 

However, this pressing setup did not necessarily account for the extra time needed to
properly melt the maleated polyolefin throughout the entire mat thickness and allow
completion of the chemical reaction between the wood and maleated polyolefin during
pressing. A study should be undertaken to optimize the pressing conditions for the one-
step panels, as these are the most likely to be industrially viable considering start-up costs
for producers that are currently making other wood-based products that are not extruded.
If the IB strength could be improved through longer pressing, perhaps at lower
temperatures or with a pre-pressing step under low pressure, these panels would be even

more advantageous from an industrial perspective.

182

APPENDIX

 

pommoonm Ho: n mam

Ems: 688.er bacon 8:662 $2.38 < 52.4 unease. sec”
6882252358728 .8. 52.4 688% as”:

 

 

 

 

 

3o 8% 3A a: a: 83% av stoned:
one 8: S: we we ovev xv 8855-3
85 84.8. in a: a: 82% av decoded:
m; can 3% a: aa 8% a: 26:85-92
2.32

OVO mNh~ o.— — m..— we ma ma awakens—Hugh: mm
m; an: on a: we oeev 8 NA:
26 cam on a: we oaev a: I:
and 32 n2 8 a: 83% we 22
23 RS 2; we a: 83% me $2
23 82 as 8 a: 82% 2 mi
23 mm: 3 _ we we 83% we 72

nJ—uhacAo—UEQA—
3:5 3% Es: aamwwa aeafiwea. pane ass 25 .25
E mo: ~52 eofiefl. 55$ has: a85.3.;

 

pfionconmm 3650 8:802 was Becca—omvam 8.: 3:688:93”: 43.22:on magnum .:.< 2an

183

 

 

 

 

 

 

3”.— moa new 90 a: v3. o 58 o .Xom
o: mwom mdm Wm mm own 0 58 N. gem
$3.2 cow—rang:
me; am: m.w_ 0.5 E N: c E8 Q .o\oom
3d 33 32 md 2 m? o 58 a .o\com
ooN 003 Wow 0m 9 wt. 0 58 n 3&8
on; omfl 5.: 5w mm o: o 58 m gem
on; omfl 52 Ex mm o: o 58 n 6&3
mad 3: WQ 2 ow N: o 58 N. .o\°m_
$6 $3 M2: vm 8 NR o 58 5 £5—
3d 83 as cm 3. N: c 58 N. .32.
omd mom 9N me 3 o: o 58 n .Xm
mm<2 toe—Ema
Ann—EV Ann—EV Ana—.6 9%“me netmmwmn< Ann—EV AEEV 25¢. .23.—
m— HOE may: 3.05—93h. .333 57.59 305—3.;

 

mg a? 230m 225 2%: _eEEcaxm é 98 €285 306282 .m.< osfl

184

 

 

 

 

 

_md 09% Sdm w mm own 0 23-0
med 23m nmdm o _m m: 0 83-6
ovd Sam No.2 fl _ m m w? o 0 ~ mmé
No.0 72% o _ .S 3 f v2. 0 me
.342 33:23::
am; mmom 0.2 c 3 NE. 0 23-0
om; chum odm n E mt. o 83-0
a; omom wm _ H M: o: o Smmé
mvd wmmm 3 S E own 0 flim—
mm<2 cue—flim—
Aam—nzv WNW—RV ”Wu—WW wamwwsm aotmmwmn< ”WWW—a“ nah—“HM"; 33‘ .053
anus—93H .533 . .

 

arm—<2 53> 950m 2053 2922 Equatoaxm Sm 3mm €3on $05382

.m.< 2an

185

200 um

 

200 um

Figure A]. SEM images of fracture surfaces of panels manufactured from maple with
maleated polypropylene, G-3003 at 250X magnification. (a) 3%, (b)
10.5% and (c) 18% MAPP.

186

100 um

100 um

(C)

100 um

 

Figure A.2. SEM images of fracture surfaces of panels manufactured from maple with
maleated polypropylene, G-3003 at 500X magnification. (a) 3%, (b)
10.5% and (c) 18% MAPP.

187

Table A.4. Mechanical Property Data for Experimental Panels Bound with 10.5%
MAPP, Pressed for 6 Minutes at 180°C and 3.4 MPa Pressure

 

 

Wood S ecies Thickness Densi MOR MOE

" 1mm) (kg/m (MPa) (MPa)
Maple 6 774 i 6 7.6 i 2 1121 i 299
Pine 6 769 i 10 7.7 i 3 1445 i 264

 

188

   

111111111111111111