Wood Modification Effects on Weathering of HDPE-Based Wood Plastic Composites

ORIGINAL PAPER

Wood Modification Effects on Weathering of HDPE-Based WoodPlastic Composites

James S. Fabiyi Æ Armando G. McDonald ÆDavid McIlroy

Published online: 18 April 2009

� Springer Science+Business Media, LLC 2009

Abstract The effects of weathering on the constituents of

wood and polymer matrix behavior in wood plastic com-

posites (WPCs) were investigated. WPCs were produced

from pine, extractives-free pine, and pine holocellulose

fibers (60%) together with HDPE (40%). These composites

were subjected to xenon-arc accelerated and outside

weathering for a total of 1200 h and 120 days, respec-

tively. The color and chemical changes that occurred on the

surface of the WPCs were analyzed using a set of analytical

techniques. For pine and extractive-free pine filled com-

posites, the results showed that the total color change,

lightness, and oxidation increased, while the lignin content

decreased. In addition, the weight average molecular

weight (Mw) and number average molecular weight (Mn)

of extracted HDPE decreased with an increase in exposure

time of the composites. However, HDPE crystallinity

increased with longer exposure time. Lightness of holo-

cellulose-based WPC changed the least while the change in

its HDPE crystallinity was not significant compared to the

other composite types. Therefore, holocellulose-based

WPC may be preferred for applications where color sta-

bility is of high priority.

Keywords Wood plastic composite � Molecular weight �Crystallinity � Wood loss � Pine � Extractive free wood �Holocellulose

Introduction

Wood plastic composites (WPC) are polymeric materials

that are made up of wood, plastic, and other chemical

additives (lubricants, coupling agents, nucleating agents,

pigments, and UV stabilizers). Incorporation of wood fibers

into plastics improves certain properties of the resulting

composite material, such as the flexural and tensile stiff-

ness, relative to pure plastics [1]. Also, the dimensional

stability of these materials tends to be greater than that of

the traditional wood products, thereby rendering them

suitable for application in end-uses where stability is a

prerequisite attribute, especially in humid and water front

environments [1]. Due to its advantages over wood and

plastic lumber, WPCs are gaining entrance into the build-

ing and construction industries [2]. However, durability

issues especially degradation and structural failure sur-

rounding the use of WPCs are of concern because of

warranty claims [3].

Indeed, several attempts have been made to improve the

durability and stability of WPC [4–8]. However, photo-

degradation has continued to be a problem with these

products. The consequences of photodegradation of WPCs

are the loss of their aesthetic appearance and strength

properties [5–7]. In order to stabilize WPC against

weathering, many approaches have been employed. These

include the incorporation of chemical additives such as

UV-stabilizers, free radical scavengers, and pigments to

moderate the effects of weathering. However, little success

has been made even with the incorporation of additives

because WPC still undergo degradation and color change

upon weathering. Stark and Matuana [6] observed that the

addition of pigments to WPC had a greater influence on

reducing the effect of photodegradation on total color

change, lightness, and mechanical properties during

J. S. Fabiyi � A. G. McDonald (&)

Forest Products Department, University of Idaho, Moscow,

ID 83844-1132, USA

e-mail: [email protected]

D. McIlroy

Physics Department, University of Idaho, Moscow,

ID 83844-0903, USA

123

J Polym Environ (2009) 17:34–48

DOI 10.1007/s10924-009-0118-y

accelerated weathering of HDPE based WPC than the

addition of UV stabilizers.

In the course of studying the durability of WPC, it was

observed that wood lignin makes a significant contribution

to the color change during weathering [4–6, 9]. Unfortu-

nately, detailed investigation on the role of lignin on color

change of weathered WPC has not been considered. The

use of pure cellulose fibers over wood fibers in WPC offers

the benefit of higher thermal stability (up to 270 �C) in that

engineering thermoplastics (e.g., nylon) can be used to

obtain materials for structural applications [10]. Therefore,

cellulose fibers (commonly produced below 200 �C during

pulping) would be a suitable reinforcement/filler for HDPE

based WPC.

Apart from the effects of weathering on the wood

component of WPC, polymer matrix degradation in WPC

during weathering can result in changes such as molecular

weight distribution (MWD) and crystallinity. Chemical

changes that occur during weathering affect the overall

properties of the polymer including melt flow/viscosity,

molecular weight, and mechanical strength [11]. Also,

reduction in molecular weight (chain-scission) leads to

shorter polymer chains and lower mechanical properties.

Torik et al., [12] stated that crystallinity is one of the most

important factors in photostability of polyethylene. Chan-

ges in the degree of crystallinity have been shown to have

pronounced effects on the mechanical behavior and frac-

ture toughness of polymers [13]. Therefore, investigating

the effects of weathering on the MWD and crystallinity of

the plastic matrix is important for outdoor structural

applications of WPC.

This study was therefore aimed at examining the

behavior of wood constituents and polymer matrix during

weathering of WPC. MWD and percent crystallinity of

polymer matrix were investigated to establish the effects of

wood modification on contribution of modified wood to

WPC surface degradation during weathering. Color and

chemical changes of modified wood fiber based WPC were

investigated to establish the role of wood extractives and

lignin during weathering.

Materials and Methods

Materials Preparation

Wood fibers were extracted in accordance with a modified

TAPPI method (T204 om-88) [14] using acetone (98%) as

solvent. One-kilogram batches of pine wood fibers (60

mesh, American Wood Fibers) was dispersed in acetone

(30 L) with constant stirring for 24 h. Thereafter, the

extract was discarded and the extracted wood was further

washed (thrice) with acetone before air dried. Note that

only approximately 3.2% extractives based on initial

weight of wood were removed by this technique. The

remaining wood fibers after the extraction would be

henceforth referred to as extractive free wood in this paper.

Holocellulose was prepared following the method

developed by Wise et al. [15]. Air-dried extractives free

wood fiber (1 kg batch) was dispersed (under constant

stirring) in 32 L of deionised water containing 0.3 kg of

NaClO2 (99%) and 200 mL of acetic acid and heated to

70 �C for 1 h. After an hour, a further aliquot of 200 mL

acetic acid and NaClO2 (0.3 kg) was added and the reac-

tion proceeded for another 1 h. This was repeated four

more times to a total of 6 h. Finally, at the completion of

the sixth treatment, the reaction was allowed to cool to

room temperature, and the delignified wood fiber was

recovered by filtering through a polypropylene screen (100

mesh). The wood fiber was washed repeatedly (7–10 times)

with deionised water until the conductivity (Accumet

conductivity meter) of the solution was reduced to 1.5–

2.5 lS. It was finally rinsed with acetone to accelerate

drying. The resultant wood fiber (holocellulose containing

1–1.5% lignin) was then air-dried. This method yielded

67% holocellulose based on initial weight of wood. Finally,

untreated pine wood flour, extractive free wood, and ho-

locellulose were dried to below 0.5% moisture content

prior to being used for WPC production.

WPC Production

HDPE (Equistar petrothene, LB 0100-00, MFI = 0.3 g/

10 min, and density = 0.950 g/cm3) and the three types of

wood fibers already prepared were used for WPC produc-

tion. One formulation based on 60% of wood and 40% of

plastic was considered. Materials were compounded and

extruded on a 35 mm counter rotating conical twin-screw

extruder (Cincinnati Milacron) to a profiled dimension of

9.5 9 38 mm. The barrel and die temperatures were set

between 149 and 193 �C. The extruded profiles were then

knife milled to a thickness of 5.0 mm for the weathering

tests because many commercial WPC products are surface

finished.

Mechanical Properties of WPC Produced

from Modified Wood Fiber

Three point flexural tests (modulus of rupture (MOR),

modulus of elasticity (MOE)) were performed in accor-

dance with ASTM Standard D 790-00 [16]. Seven repli-

cates were performed for each WPC type (20 9

5.73 9 115 mm) on an Instron 5500R Universal test

machine equipped with a 454 kg load cell. Data was col-

lected and processed using Bluehill software (Instron).

J Polym Environ (2009) 17:34–48 35

123

Accelerated and Natural Weathering of WPC

WPC specimens (5 9 38 9 101 mm) were subjected to

accelerated weathering test, which was conducted in a

xenon-arc weatherometer (Q-Sun). The average irradiance

was 0.70 W/m2 at 340 nm wavelength with a chamber

temperature of 70 �C and water spray in accordance with

ASTM D 6662 [17], which is a very severe weathering

condition. The natural weathering test was conducted by

exposing the WPC specimens (5 9 38 9 610 mm) outside

in Moscow, Idaho, USA on a south-facing wall at an angle

of 45� (Moscow, ID) in accordance with ASTM D 1435

[18]. Average daylight in Moscow within the period of

exterior exposure was approximately 12–14 h/day for July

to August 2006 and 9–10 h/day for September to

November 2006. The sample condition was assessed from

0 to 1200 h and 0 to 120 days of exposure in xenon-arc and

outside weathering regimes, respectively.

Material Characterization after Weathering

Scanning Electron Microscopy Analysis

Scanning electron micrographs of WPC samples were

obtained on a LEO Gemini field emission SEM instrument.

Sections of unweathered and weathered (400 and 1200 h

xenon-arc weathered) WPCs were cut into approximately

8 9 8 mm pieces. The specimens were mounted on alu-

minum stubs using carbon tape. The weathered surfaces

were analyzed directly (without coating) at 1 kV at a

magnification of 709.

Color Measurement

A StellarNet EPP2000 UV-Vis spectrometer (190–

850 nm) using a krypton light source (SL1, Stellar Net)

and a diffuse reflection fiber optic probe was used to

measure color in accordance with the ASTM 2244 stan-

dard [19]. The spectrometer SpectraWiz software trans-

forms spectral data into CIELAB color coordinates based

on a D65 light source (L*, a* and b*) [20]. Color was

measured for five replicates per WPC sample at three

locations on each specimen. Relative lightness (DLrel) and

total color change (DEab) were calculated using the fol-

lowing equations:

DLrel ¼L�final � L�initial

L�initial

� 100 ð1Þ

DEab ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

DL2 þ Da2 þ Db2p

ð2Þ

where, DL, Da, and Db represent the differences between

the initial and final values of L*, a*, and b*, respectively.

An increase in L means the sample is lightening.

Infrared Spectroscopy Analysis

Infrared measurements were performed with a Thermo-

Nicolet Avatar 370 FTIR spectrometer operating in the

attenuated total reflection (ATR) mode (SmartPerformer,

ZnSe crystal). Thin slices of about 50 lm of WPC (vacuum

dried) from three samples were analyzed on the exposed

side. Each spectrum was taken as an average of 64 scans at

a resolution of 4 cm-1. Special interest was focused on the

C=C double bond region (1630–1660 cm-1), carbonyl

region (1660–1800 cm-1), and the hydroxyl region (3200–

3500 cm-1) [21–23].

For the quantitative analysis, the spectra were normal-

ized and the absorbance of the peak of interest was

obtained. The concentrations of the carbonyl (C=O),

unsaturated carbon (C=C), and hydroxyl (OH) groups

(degradation products) were estimated based on the Lam-

bert–Beer equation (Eq. 3):

A ¼ ebc ð3Þ

where, A is the absorbance of the functional group band in

the infrared spectrum, c is the molar concentration of the

functional group mol/L (M), e is the molar absorptivity

(L/mol/cm), and b is the path length of the sample (that is,

the optical path of the infrared beam through the sample,

cm). A is obtained from the absorption band of interest in

the ATR spectrum. Values for the molar absorptivity, e used

in this study were taken from the work reported by Lacoste

et al., [24] from the spectra of model carboxylic acids

(1718 cm-1), esters (1744 cm-1), and C=C (1635 cm-1),

which were 350, 590 and 121 L/mol/cm, respectively. The

molar absorptivity of hydroperoxide from the secondary

alcohol was 660 L/mol/cm [23], this was used for the

investigation of weathering on wood. Optical path length b

was considered as the product of the effective path length,

de and the number of reflections in the internal reflection

element (IRE). The protocol of Fabiyi et al. [25] was

employed to determine the concentration of each functional

group was calculated as:

c ¼ A�

2dpe� �

ð4Þ

where de = 2dp and dp is the ATR depth of penetration.

X-ray Photoelectron Spectroscopy (XPS) Analysis

X-ray photoelectron spectroscopy (XPS) is a non-destruc-

tive surface analytical technique that provides information

on the oxidation or chemical bonding state of elements [26,

27]. It has been used for the characterization of wood fibers

[28] and the weathering of wood and WPC [27].

XPS measurements were performed on a limited number

of samples on both control and weathered WPC (made

from HDPE and untreated pine wood fibers) to determine

36 J Polym Environ (2009) 17:34–48

123

surface chemical composition. A custom built spectrometer

using two monochromators at 1253 and 1487 eV, with a

resolution of 50 meV (Courtesy of Dr. D. McIlroy Labo-

ratory, Dept of Physics, Univ. of Idaho) was employed. The

instrument has a high-resolution hemispherical analyzer.

Charging was minimized using a flood gun. Samples were

mounted on a stainless steel sample holder with carbon

tape. Care was taken to ensure that uncontaminated sample

with flat surface covered the carbon tape to avoid its

(carbon tape) exposure to X-ray beam.

Two types of spectra were obtained that include low

resolution spectra from 0 to 1100 eV binding energy to

determine elemental composition (oxygen and carbon

ratios) and high resolution spectra from 280 to 300 eV to

analyze carbon. For the quantitative analysis, the spectra

were baseline corrected, normalized, and curve fit using

IGOR Pro 5.05 software (WaveMetrics, Inc). The area of

each peak identified by curve fitting was mathematically

computed.

The following data processing restrictions were made in

accordance with procedure developed by McDonald et al.

[28]:

(i) The full widths at half height (FWHM), for the C1 to

4 peaks (285, 286.5, 288, and 289.5 eV) were kept

constant (1.6 eV).

(ii) The fitted peaks was made using Voigt.

(iii) The third restriction is that a peak C0 (C=C), at

283.5 eV was fitted when the residual of the band

indicating the presence of this peak (FWHM of 1.6

EV).

(iv) Therefore, with these restrictions in place, a peak

fitting routine was then used to minimize and

randomize the residual signal.

From the area of the curve fitted peaks, the percent

carbon, oxygen, and the ratios of oxygen to carbon (O/C)

present upon weathering were computed.

Pyrolysis Gas Chromatography-Mass Spectrometry

Wood derived compounds (lignin and polysaccharides)

were identified and quantified by GC-MS according to the

method described by Fabiyi et al., [25] and Schauwecker

et al., [29]. Two replicates specimens were used for each

treatment. Wood content in the weathered WPC samples

was quantified using pyrolysis-GC-MS by developing a

calibration curve from the total peak areas under the wood

derived peaks relative to the total peak areas under the

polyethylene derived peaks. Calibration curves were based

on tests of a series of WPC formulations of known wood

content (untreated and extractive free pine) ranging from 0

to 100% in HDPE-based WPC.

Polymer Matrix Characterization

The need to extract the polymer matrix (HDPE) prior to

analysis is important since one of the main interests was to

understand its behavior. The WPC surface (unweathered

and weathered) was scrapped with a razor blade and HDPE

was extracted. Approximately 500 mg of surface material

was weighed into a flat bottom flask and 1,2,4-trichloro-

benzene (TCB) (50 mL) was added. The contents were

then rapidly heated with a microwave oven (Sanyo,

600 W) for 2 to 3 min to minimize oxidation. The solu-

bilized HDPE was collected by separating wood fibers

from the heated contents using a cotton wool filter, then re-

filtered through filter paper (Whatman No.1) and dried

under a stream of nitrogen to obtain a film of HDPE.

Polymer Analysis by Gel Permeation Chromatography

The effect of weathering on the molecular weight and

MWD of plastic (HDPE) was determined by gel perme-

ation chromatography (GPC) on elution with TCB. Mw and

Mn for HDPE was conducted by Equistar Chemicals

(Cincinnati, OH) and ExxonMobil Chemicals (Baton

Rouge, LA) using high temperature GPC systems.

The HDPE from xenon-arc weathered extractive free

and exterior weathered pine composites were analyzed by

Equistar Chemicals. The HDPE were heated for 1 h at

175 �C in TCB (1–1.5 mg/mL) prior to injection (300 lL).

Analysis was performed on a Waters GPC2000 system and

separation was achieved using two mixed-bed columns at

145 �C and a TCB flow rate of 1 mL/min. The components

were detected using refractive index (RI), capillary vis-

cometer, and Precision Detector LS systems. The GPC

system was calibrated using a universal calibration curve

(PS standard, K = 0.0001387, a = 0.70) with a linear fit.

The extracted HDPE from unweathered and xenon-arc

weathered pine composites were also analyzed by Exxon-

Mobil Chemicals. Analysis was performed on GPC 220

system with an autosampler equipped with an RI and dif-

ferential viscosity detector (Viscotek). Separation was

carried out using three columns in series (10 lm mixed-

bed, 300 mm 9 7.6 mm from Polymer Laboratories). The

analysis was performed at 135 �C using TCB as the mobile

phase (1 mL/min). Calibration was conducted using a

narrow range PS standard. Molecular weight analysis was

performed using OmniSEC 4.1 software (Viscotek).

Polymer Crystallinity using Differential Scanning

Calorimetric

Differential scanning calorimetric (DSC) was performed on

a TA2910 instrument to monitor the thermal behavior of

HDPE crystallization according to ASTM D 3418 [30].

J Polym Environ (2009) 17:34–48 37

123

The instrument was calibrated using indium. TCB-extrac-

ted HDPE (5–7 mg) was analyzed using a temperature

profile after 2 min equilibration at 50 �C, then ramped to

160 �C at a heating rate of 10 �C/min and held isother-

mally for 10 min, cooled to 50 �C at -10 �C/min. Two

replicate measurements were performed for each sample.

The degree of crystallinity was calculated from the

experimentally determined heat of fusion for the extracted

HDPE and the heat of fusion corresponding to a pure PE

crystal of 293.6 J/g [31, 32].

All the data was statistically analyzed using SPSS v11

software.

Results and Discussion

Flexural Properties of Modified Wood Based WPC

Flexural properties of modified wood based WPC are

presented in Table 1. The results showed that there were no

significant differences in the MOR, and MOE among the

composites made from pine fibers, extractive free pine

fibers, and pine holocellulose fibers. This was contrary to

the expectation because holocellulose based composites

supposed to be more hydrophilic (incompatible) than pine

fibers and extractive free pine fibers.

Characterization of WPC Weathered Surface

Visual Appearance

Figure 1 shows scanning electron micrographs of unweath-

ered, 400, and 1200 h xenon-arc weathered composites.

These composites degraded differently depending on wood

fiber types. Cracks were observed on the surface, which

might be as an effect of polymer chain-scission that normally

results in highly crystallized polymer zones. Cracking occurs

because of wetting and drying cycles between the surface

and interior sections. The extent of wood degradation and

erosion, leaving cracks and ‘‘pits’’ increased upon extended

weathering. These findings are in agreement with other

studies [33, 34]. Also, more wood fibers were found on the

surface of holocellulose based composites upon longer

exposure compared to pine based composites. After 400 h of

exposure in xenon-arc weatherometer, the surface layer of

the weathered WPC was eroded; thereby creating cavities on

the surface (Fig. 1b and e). Accelerated weathering exposure

for 1200 h resulted in the size and frequency of cavities

(Fig. 1c and f).

Color Changes

Relative lightness (DLrel) and total color change (DEab) of

xenon-arc and exterior weathered WPC made from

untreated, extractive free and holocellulose are shown in

Figs. 2 and 3. These results showed that WPC produced

from holocellulose fibers had the lowest DLrel and DEab for

both xenon-arc and outside weathering regimes. Stark [35]

conducted a study on the color change of HDPE/pine

composites in which HDPE (48.5 or 47%), pine flour

(50%), UV absorber (0.5 or 1%) and zinc ferrite pigment

(1 or 2%) were used. Surprisingly, HDPE/holocellulose

composites (after 1200 h xenon-arc exposure) performed

better in term of lightness than HDPE/pine composites with

UV absorber (0.5 or 1%) and zinc ferrite pigment (1 or 2%)

after 1000 h of xenon arc weathering.

There was no significant difference in DLrel for the WPC

produced from untreated and extractive free pine wood

(a = 0.05) subjected to both xenon arc and outside

weathering. This result is expected since the extractives

were not considered a significant contributor to weathering.

The DLrel and DEab for all the WPC increased upon

weathering until 1200 h and 120 days for xenon-arc and

outside weathering regimes, respectively (Figs. 2 and 3).

There was no significant difference between DEab of

holocellulose and untreated/extractive free composites.

This might be due to the contribution of combined effect of

the three color coordinates. The results obtained were

consistent with other weathering studies [4, 5].

Surface Chemical Characterization

IR Spectroscopy

IR spectrum of WPC showing the bands between carbonyl

group regions is presented in Fig. 4. The entire IR spectra

(not shown) conveyed some useful information on some

bands. For instance, the spectral structures between 1015–

1050 and 3500–3080 cm-1 regions which were assigned

to C–O and OH groups in wood (combination of cellulose,

hemicellulose and lignin) decreased upon weathering of the

WPC [36, 37]. It was shown that the lignin ether linkage

assigned band (1512–1508 cm-1) was not present in the

holocellulose composites. This implies that the lignin

content in the holocellulose was negligible. For pine and

Table 1 Flexural properties of modified wood fiber/HDPE

composites

WPC type MOR (MPa) MOE (GPa)

Pine 25.51 (2.5) 2.25 (3.2)

Extractive free pine 25.57 (5.0) 2.41 (7.4)

Holocellulose 25.03 (3.9) 2.54 (3.1)

Data presented are the average of seven replicated specimens while

those in bracket are the coefficients of variation

38 J Polym Environ (2009) 17:34–48

123

extractive free pine WPC, the band at 1508–1512 cm-1

was shown to decrease in intensity as a function of

weathering time [37]. This indicates that lignin degradation

occurred on the weathered WPC surface.

In this study, the extent of WPC oxidation was deter-

mined by its carbonyl functionality (1750–1690 cm-1).

The concentration of the carbonyl (carboxylic acids at

1715 cm-1 and esters at 1735 cm-1) increased upon WPC

weathering (Figs. 5 and 6). The increase in carbonyl con-

centration provided evidence that photodegradation

occurred with prolonged exposure time. It also means that

the composites were vulnerable to further degradation

because these carbonyl groups are photolabile [4]. The

concentration of the C=C groups at 1635 cm-1 present in

weathered WPC slightly increased upon weathering

(Fig. 7). This suggests that HDPE and wood oxidation

occurred resulting in an increase unsaturated functionality.

In addition, the hydroperoxide (from hydroxyl group at

3500–3080 cm-1) concentration decreased with longer

exposure time (Fig. 8). This indicates that wood content

decreased from the weathered surface since the band

between 3500 and 3080 cm-1 are usually assigned to pri-

mary alcohol. However, wood loss from holocellulose-

based WPC was minimal as compared to pine and

extractives free pine based composites.

X-ray Photoelectron Spectroscopy (XPS)

The results obtained from XPS analysis (widescans) for the

unweathered and xenon-arc weathered HDPE/pine com-

posites showed only the presence of carbon and oxygen.

The peaks for C1s and O1s had chemical binding energies

of approximately 285–290 and 532 eV, respectively,

(Table 2) [38, 39]. High resolution scans of the C1s region

were analyzed after curve fitting and the spectra shown in

Fig. 9, while binding energies are given in Table 2. The

Fig. 1 Scanning electron

micrographs (709) of HDPE-

pine composites a unweathered,

b 400 h xenon-arc weathered,

c 1200 h xenon-arc weathered,

and HDPE-holocellulose

composites d unweathered,

e 400 h xenon-arc weathered,

f 1200 h xenon-arc weathered

J Polym Environ (2009) 17:34–48 39

123

presence of C0 and C1 in the composites may be attributed

to the aliphatic and aromatic carbons in lignin and

extractives and/or HDPE matrix. Freudenberg and Neish

[40] outlined some assumptions relating to the origin of C1s

carbons in wood and these were adopted in this study.

These assumptions are thereby outlined. In ideal situations,

cellulose lacks C1 because of the polysaccharide structure;

however, 49% of the carbon atoms consist of C1 type in

milled wood lignin. Furthermore, C2 could be attributed to

primary and secondary alcohols in lignin, extractives, and

polysaccharides. The presence of C3 carbon could be

associated with the aliphatic or aromatic ethers in lignin

and extractives, and to carbonyl groups (carboxylic,

ketones, esters or aldehydes) in lignin, extractives, and

hemicellulose. Also, C4 could be a contribution from the

carboxylic acids and esters in lignin, extractives and cel-

lulose. In addition, C3 and C4 may also be an indication of

carbonyl groups formation at the weathered WPC surface.

This is because the carbonyl concentrations increased upon

weathering as observed under IR analysis. Lojewska, et al.,

[41] stated that the carbon atoms, which are most prone to

oxidation, occupy the 2, 3, and 6 positions in glycopyr-

anosyl moieties of the cellulose chain.

Figure 9 showed that there was a significant increase in

C3 and C4 peak from 0 to 1200 h of WPC exposure in

xenon-arc weathering regime. Table 3 show that C0 (C=C)

10

20

30

40

50

12008004000

Ligh

tnes

s ( ∆

Lre

l, %

)

Exposure time (h)

Pine

Extractive free

Holocellulose

5

15

25

35

Col

or c

hang

e (∆

Eab

)

PineExtractive freeHolocellulose

12008004000

Exposure time (h)

(a)

(b)

Fig. 2 Effect of xenon-arc weathering on the a DLrel and b DEab of

HDPE based WPC made from pine, extractives free pine, and pine

holocellulose

-5

5

15

0 30 60 90 120Sur

face

ligh

tnes

s (

L rel, %

)

Exposure time (days)

PineExtractive freeHolocellulose

0

5

10

15

Col

or c

hang

e (

Eab

)PineExtractive freeHolocellulose

∆∆

0 30 60 90 120


(a)

(b)

Fig. 3 Effect of outside weathering on the a DLrel and b DEab of

WPC made from pine, extractives free pine, and pine holocellulose

0.8

0.7

0.6

0.5

0.4

0.3

1750 1700 1650

Carboxylic acid

1200h

Abs

orba

nce

Wavenumber (cm-1)

Ester

Fig. 4 Curve fitted for IR carbonyl region (1800–1680 cm-1) of

1200 xenon-arc weathered HDPE-pine composite

40 J Polym Environ (2009) 17:34–48

123

and C1 (C–C or C–H) peak intensity decreased upon

weathering, indicating a decrease in C1 peak. This is con-

trary to the result obtained by IR data. The decreased C=C

groups from XPS data may be due to the contribution of the

decreased lignin content at the surface of weathered WPC.

On the other hand, C2, C3, and C4, which correspond to

C–OH, O–C–O or C=O, and O–C=O respectively,

increased upon WPC weathering. This suggests that the

concentrations of the oxidized carbons increase with longer

exposure time of composites to varying environmental

conditions. Note that C3 and C4 contributed mostly to the

increase in surface oxidation relative to C2 as they

increased from 8 to 21% (C3) and 7 to 17% and (C4)

(Table 3). Indeed, C2, C3 and C4 increased by 38, 167 and

153%, respectively, comparatively to the unweathered

composites. Relative atomic weight (%) for carbon and

oxygen were obtained from the integration of peak areas of

the different emissions corrected for intensity with an

appropriate sensitivity factor for each of the XPS spectra.

The carbon and oxygen compositions are presented in

Table 3. In order to substantiate that surface oxidation had

occurred, as suggested from the increased oxidized car-

bons, the degree of surface oxidation was calculated by

finding the ratios of oxygen to carbon atoms. In addition,

the relationship (Eq. 5) developed by Matuana and Kam-

dem [27] for the computation of the ratio of oxidized to

unoxidized carbon atoms was adopted.

Coxidized

Cunoxidized

¼ C2þ C3þ C4

C1ð5Þ

The ratios of oxygen to carbon and oxidized to

unoxidized carbon atoms increased upon weathering

(Fig. 10 and Table 3). This is an indication that oxidation

had occurred. This provides supporting evidence to the

increased carbonyl groups concentrations observed by IR

analysis. Matuana and Kamdem [27] suggested that the

0 400 800 12000.0

0.4

0.8

1.2

1.6

2.0

Pine Extractive free Holocellulose

Con

cent

ratio

n at

171

5 cm

-1 (

mol

/kg)

Exposure time (h)

0.4

0.8


Con

cent

ratio

n at

173

5 cm

-1 (

mol

/kg)

0 400 800 1200Exposure time (h)

(a)

(b)

Fig. 5 Effect of xenon-arc accelerated weathering on the concentra-

tion of a carboxylic acid, 1715 cm-1 and b esters, 1735 cm-1 present

in composites made from pine, extractives free pine, and pine

holocellulose

0 30 60 90 1200.0

0.4

0.8

1.2

1.6

2.0

2.4


Con

cent

ratio

n at

171

5 cm

-1 (

mol

/kg)


0.0

0.4

0.8

1.2

1.6


Con

cent

ratio

n at

173

5 cm

-1 (

mol

/kg)

0 30 60 90 120Exposure time (days)

(a)

(b)

Fig. 6 Effect of outside weathering on the concentration of a carbox-

ylic acid, 1715 cm-1 and b esters, 1735 cm-1 present in composites

made from pine, extractives free pine, and pine holocellulose

J Polym Environ (2009) 17:34–48 41

123

increased oxidation due to weathering might probably

have occurred from the oxidized carbon, such as C=O and

O–C=O. This might have originated from the carbonyl

functional groups of wood fiber.

WPC Material Composition and Characterization

Pyrolysis Gas Chromatography-Mass Spectrometry

The chromatograms of the pyrolyzed unweathered and

weathered composites produced from untreated and

extractive free pine wood fibers are shown in Figs. 11 and

12 respectively. Molecular fragments from both wood and

plastic were obtained from the pyrolyzed specimens. The

wood derived compounds in HDPE/pine composites are

given in Table 4 [9, 42]. The same wood derived peaks

were found in extractive free wood fibers composites with

some variation in quantity.

Semi-quantitative analysis was conducted based on the

calibration technique employed. The total percentage area

of all the wood derived compounds identified at the surface

0 400 800 12000

1

2

3

4C

once

ntra

tion

at 1

635

cm-1 (

mol

/kg)

Exposure time (h)


0 30 60 90 1200

1

2

3

4

Con

cent

ratio

n at

163

5 cm

-1 (

mol

/kg)



(a)

(b)

Fig. 7 Effect of a xenon-arc and b outside weathering on the C=C

(1635 cm-1) concentration in HDPE-based WPC made from pine,

extractives free pine, and pine holocellulose

0 400 800 12000

1

2

3

4

5

6

7

Con

cent

ratio

n at

334

0 cm

-1 (

mol

/kg)

Exposure time (h)


0 30 60 90 1200

2

4

6

8

Con

cent

ratio

n at

334

0 cm

-1 (

mol

/kg)



(a)

(b)

Fig. 8 Effect of a xenon-arc and b outside weathering on the

hydroxyl (3340 cm-1) concentration in HDPE-based WPC made

from pine, extractives free pine, and pine holocellulose

Table 2 Assignments of peaks with their corresponding binding

energy and bond type for XPS scan of the C1s and O1s regions of

xenon-arc weathered HDPE/pine composites

Element group Binding energy (eV) Bond type

C1s C0 282.5 C=C

C1 285.0 C–C or C–H

C2 286.5 C–OH

C3 288.0 O–C–O or C=O

C4 289.5 O–C=O

O1s O1 530.4

O2 532.3

O3 535.6

Source: Dorris and Gray [38]; Grigsby et al. [39]

42 J Polym Environ (2009) 17:34–48

123

of unweathered and weathered WPC was used to account

for the wood content. Figure 13a shows that the wood

content at the WPC surface decreased upon xenon-arc

weathering from 60 to 38% and 60 to 39% for pine and

extractive free wood composites, respectively. These

results indicate that approximately 36% of the original

wood content in the WPC was lost after 1200 h of xenon-

arc weathering. Wood loss between untreated and extrac-

tive free wood composites was significant (a = 0.05). This

is expected because color measurement showed that there

was no significant change in the visual appearance between

both composites upon weathering.

Wood content of the exterior exposed WPC also

decreased with longer exposure time (Fig. 13b). The results

showed that for pine and extractive free wood composites,

wood decreased from 60 to 41% and 60 to 46%, respec-

tively. This indicates that about 32% and 27% of original

wood content in untreated and extractive free wood based

WPC, respectively were lost from the surface due to its

120 days of exterior exposure. Also, there was no signifi-

cant difference in the quantity of degraded wood

(a = 0.05) between untreated and extractive free wood

composites.

1.0

0.8

0.6

0.4

0.2

0.0

Inte

nsity

(x1

05)

300 295 290 285 280

Binding energy (eV)

C0

C1

C2

C3C4

Unweathered

1.0

0.8

0.6

0.4

0.2

0.0

300 295 290 285 280Binding energy (eV)

C0

C1

C2

C3

C4

1200 h

Inte

nsity

(x1

05)

Fig. 9 High resolution XPS spectra and result of curve fitting of C1 s

region of unweathered (top) and xenon-arc weathered (bottom)

HDPE/pine composites

Table 3 Surface composition of carbon and oxygen of xenon-arc

weathered untreated pine composites

Peak (%) Exposure time (h)

0 100 200 400 800 1200

C0 5.9 2.8 2.4 1.3 1.0 1.1

C1 59.4 48.8 42.0 40.4 38.7 40.1

C2 20.2 21.4 27.5 27.9 26.5 24.3

C3 7.8 17.6 18.5 20.6 20.8 17.7

C4 6.7 9.4 9.6 9.8 12.9 16.9

Carbon 63.9 60.1 58.9 55.8 55.8 52.3

Oxygen 36.1 39.9 41.1 44.2 44.2 47.7

The data presented were the average of two replicates while their

coefficients of variation were below 8%

0 400 800 12000.0

0.4

0.8

1.2

1.6

O/C

Surface oxidation

Rel

ativ

e ab

unda

nce

Exposure time (h)

Fig. 10 Effect of xenon-arc weathering on the oxygen to carbon and

oxidized-to-unoxidized carbon atoms ratios determined by XPS on

weathered HDPE/pine composites surface

0 5 10 15 20 25 30 35 40

*** * **

** *

**

**** ** *

0 5 10 15 20 25 30 35 40Time (min)

** *

1200h

0h HDPE/pine* Lignin derived

Fig. 11 Py-GC-MS chromatograms of (top) unweathered and (bot-tom) xenon-arc weathered pine composites after 1200 h

J Polym Environ (2009) 17:34–48 43

123

These data (i.e., from pine and extractive free pine fibers

based composites) suggest that the wood was preferentially

degraded during WPC weathering. This conclusion is

supported by SEM observations and IR spectral analysis.

Comparing this observation with that of SEM, colorimetry,

and IR, it was evident, that lignin is more sensitive to

photodegradation than holocellulose (cellulose and hemi-

celluloses). In this study, the use of holocellulose improved

the weathering performance (color stability and wood

retention) of WPC without any additives and coupling

agents.

Polymer Matrix Characterization

Molecular Weight Distribution of the Weathered Polymer

Matrix

GPC chromatograms of extracted HDPE are shown in

Fig. 14. This showed that Mw and peak intensity decreased

upon weathering. Also, the presence of multiple peaks at

lower molecular weight was observed because of longer

1200 h xenon-arc weathered

0h HDPE/extractive free* Lignin derived

5 10 15 20 25 30 35 40Time (min)

0

5 10 15 20 25 30 35 400

**

** *

* ** * ** *

**

**

**

*

*

Fig. 12 Py-GC-MS chromatograms of (top) unweathered and (bot-tom) xenon-arc weathered extractive free wood composites after

1200 h

Table 4 Wood derived compounds in pyrolyzed HDPE/pine WPC

Retention time (min) Compound

1.27 Carbondioxide

4.87 3-furancarboaldehyde

5.17 5-methyl-2-furancarboaldehyde

6.30 Phenol

6.59 Benzaldehyde

7.51 2,3,5-trimethylfuran

7.90 2-methoxyfuran

8.98 Guaiacol

8.97 1,2-furanlethanone

9.52 Acetophenone

10.43 4-hydroxy-3-methylbenzaldehyde

10.96 2(methoxymethyl)-furan

11.88 Methylhydroquinone

12.23 2-methyl-phenol

13.00 3,4-dimethylphenol

15.01 1-methoxy-3-methybenzene

15.67 4-methylguaiacol

16.48 Hydroquinone

17.51 3-methyl-4-ethylphenol

18.53 Isovanillin

19.40 4-vinylguaiacol

20.56 2-methoxy-4(1-propenyl-phenol)

21.87 Isoeugenol

22.71 4-propylguaiacol

24.54 Guaiacylacetone

26.40 4-propenyl-syringol

0 400 800 120035

40

45

50

55

60

Pine

Extractive free

Rel

ativ

e ab

unda

nce

(%)

Exposure time (h)

0 30 60 90 12035

40

45

50

55

60

Rel

ativ

e ab

unda

nce

(%)


Pine Extractive free

(a)

(b)

Fig. 13 Wood content of a xenon-arc and b outside weathered WPC

44 J Polym Environ (2009) 17:34–48

123

exposure time. Multiples peaks were observable starting

from 400 h xenon-arc weathered but not found in the

exterior weathered (120 days) extracted HDPE. This trend

of polymer scission is consistent with other report [43].

From Figs. 15 and 16, Mw and Mn decreased upon extended

exposure time for both xenon-arc and outside weathering of

WPC. However, there is an exception to Mw of HDPE from

pine wood composites with no significant difference

throughout the exposure time in xenon-arc (a = 0.05).

It was reported that chain scission reactions have a

direct influence on Mn, and that the number of apparent

chain scissions per molecule, Nscission, can be obtained from

a relationship as expressed in Eq. 6 [44]

Nscission ¼Mn;0

Mn;t� 1 ð6Þ

where Mn,0 and Mn,t are the number average molecular

weights of the initial (unweathered) and subsequence

exposure time, respectively.

Another approach for computing chain scission reac-

tions is expressed in terms of the number of scission events

per g of such material as given in Eq. 7 [45]:

Nt ¼1

Mn;t� 1

Mn;0ð7Þ

From Fig. 17, the number of chain scissions (both in

scission/mL and scission events/g) increased with longer

exposure time of the WPC in outside and xenon-arc

weathering regimes. The relationship between the number

of scissions and exposure time is linear up to 90 days

(outside weathering). This implies that degradation process

occurs with random scissions [46]. In addition, the number

of chain scissions that took place in HDPE began to

decrease after 800 h of exposure in xenon-arc weathering

regime. This suggests that crosslinking was becoming more

predominant upon longer exposure of HDPE-based WPC.

Polymer Matrix Crystallinity

Figure 18 shows a DSC thermogram of extracted HDPE

from holocellulose composites. The percent crystallinity of

HDPE in xenon-arc and exterior weathered untreated pine

wood and holocellulose composites were determined and

shown in Fig. 19. The extractive free wood composites was

not considered for this analysis because pyrolysis GC-MS

0.0

0.2

0.4

0.6

0.8

1.0

Slice Log MW2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0

0h

400h

1200h

dwt/d

(logM

)dw

t/d(lo

gM)

0.0

0.2

0.4

0.6

0.8

1.0

Slice Log MW2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0

0 day

60 days

90 days

120 days

(a)

(b)

Fig. 14 Gel permeation chromatograms of extracted HDPE from axenon-arc and b outside weathered HDPE/pine composites in 1,2,4-

trichlorobenzene

0

50000

100000

150000

200000

250000

0 400 800 1200

Exposure time (h)

Mw

0

5000

10000

15000

20000

25000

30000

Mw

Pine Extractive

Pine Extractive

0 400 800 1200

Exposure time (h)

(a)

(b)

Fig. 15 Effect of xenon-arc weathering on a Mw and b Mn of

extracted HDPE from pine and extractive free pine wood fibers

composites

J Polym Environ (2009) 17:34–48 45

123

results showed that there was little or no difference

between the untreated and extractive free wood based

WPC.

The change in HDPE crystallinity was not significant in

holocellulose-based WPC upon weathering. The HDPE in

weathered untreated pine composites experienced an

increase in percent crystallinity during xenon-arc and out-

side weathering (Fig. 19). These observations agree with

the report of other studies [4, 47]. Jabarin and Lofgren [47]

reported that weathering of HDPE resulted in increased

polymer crystallinity. Stark and Matuana [9] stated that the

increase in crystallinity is an indicator of PE chain scission

during photodegradation. Increased crystallinity may be

due to the changes in molecular weight, which occurred

because of polymer degradation that caused chain breaking

and probably secondary crystallization. In addition,

increase in crystallinity is due to decrease in Mw of the

polymer [47].

Indeed, further degradation may be initiated by the free

radicals and short chain molecules generated during

weathering. It was explained that chain scissions resulting

from degradation via Norrish I and II reactions reduces the

density of entanglements in the amorphous phase [34].

Consequently, crystallization of shorter molecules occurs

due to higher mobility. Shorter chains have higher mobility

and can easily crystallize thereby resulting in an increase in

crystallinity [47]. This chemi-crystallization could be evi-

dent by the cracks that occurred at the weathered com-

posites’ surface. Severe cracking causes embrittlement and

thus negatively influences the mechanical properties [34].

Since change in HDPE crystallinity was very small in

holocellulose-based WPC, it may mean that holocellulose

play an important role in hindering the reduction of density

of entangled amorphous phase of the plastic.

Conclusions

The changes that occur to wood and plastic components of

WPC (produced from modified wood fibers—extractives

0

50000

100000

150000

200000

250000

0 60 90 120

Exposure time (day)

Mw

0

5000

10000

15000

20000

25000

Mn

0 60 90 120

Exposure time (day)

(a)

(b)

Fig. 16 Effect of outside weathering on a Mw and b Mn of extracted

HDPE from pine wood fibers composites

0 400 800 12000

3

6

9

12

Num

ber

of c

hain

sci

ssio

n

Exposure time (h)

Pine (scissions/g) Pine {(scissions/ml) x10-5} Extractive free (scissions/g) Extractive free {(scissions/ml) x10-5}

0 30 60 90 1200

3

6

9

12

15

18

Num

ber

of c

hain

sci

ssio

n


Pine (scissions/g) Pine {(scissions/ml) x10-5}

(a)

(b)

Fig. 17 Apparent number of chain scissions in the oxidation reaction

of a xenon-arc and b outside weathered WPC

46 J Polym Environ (2009) 17:34–48

123

and lignin removal) during natural and accelerated

weathering were investigated. SEM showed that WPC

produced from the modified wood weathered differently. It

also showed that the degree of wood degradation and

erosion, leaving cracks and ‘‘pits’’ increased upon extended

weathering. Interestingly, more wood fibers were found on

the surface of holocellulose composites upon longer

exposure compared to other composites (pine and extrac-

tive free wood fibers). Weathering caused increase in total

color change and lightness. It was found that WPC pro-

duced from holocellulose fibers had the lowest total color

change and lightness for both outside and xenon-arc

weathering regimes. The extent of WPC oxidation on the

weathered surface was determined by its carbonyl func-

tionality’s concentration, which increased upon longer

exposure time. The increase in carbonyl concentration

provided evidence that photodegradation occurred with

prolonged exposure time. Wood was lost during weath-

ering of pine and extractive free wood based WPC. The

general trend of WPC weathering shows that the polymer

molecular weight decreased while the crystallinity

increased upon longer exposure time. From this study,

change in HDPE crystallinity was not significant in holo-

cellulose composites for the duration of exposure (1200 h).

The HDPE in weathered pine composites experienced an

increased in percent crystallinity during both xenon-arc and

outside weathering. In addition, increase in crystallinity

may be due to decrease in Mw of the polymer. From the

results obtained by different analytical techniques, it was

evident that lignin is more sensitive to photodegradation

than holocellulose (cellulose and hemicelluloses). On the

other hand, the use of holocellulose improved the weath-

ering performance (color stability and wood retention) of

WPC without any additives and coupling agents. There-

fore, holocellulose-based WPC may be preferred for

applications where color stability is of high priority.

However, dimensional instability and biodegradation may

be associated problems with the use of holocellulose

composites (especially white rot fungi).

Acknowledgment The authors sincerely acknowledge (i) funding

by a grant from the USDA Forest Products Laboratory, (ii) the

National Research Initiative of the USDA Cooperative State

Research, Education and Extension Service (grant number 2005-

35103-15243) provided support for the FT-IR spectrometer, (iii) Drs.

Michael Wolcott and Karl Englund for technical assistance, and (iv)

Gary Stark and Clay Enos for GPC analysis at Equistar Company, and

Tony Poloso for GPC analysis at ExxonMobil Chemicals.

40 60 80 100 120 140 160

-2

0

2

4

Unweathered

1200 h

Hea

t flo

w (

W/g

)

Temperature (°C)Exo Up

Fig. 18 DSC thermograms of extracted HDPE from unweathered and

xenon-arc weathered holocellulose composites (bottom curves are

from the heating cycle while the top curves are from the cooling

cycle)

0 400 800 1200

40

50

Cry

stal

linity

(%

)

Exposure time (h)

Pine

Holocellulose

0 30 60 90 12035

40

45

50

Pine Holocellulose

Cry

stal

linity

(%

)


(a)

(b)

Fig. 19 Effects of a xenon-arc and b outside weathering on the

crystallinity of extracted HDPE from weathered WPC made from

pine and holocellulose fibers

J Polym Environ (2009) 17:34–48 47

123

References

1. Clemons C (2002) For Prod J 52:10–18

2. Smith PM, Wolcott MP (2005) In: Proceedings of the 8th inter-

national conference on woodfiber-plastic composites, Madison,

WI, pp 335–343, May 23–25

3. Klyosov A (2006) In: Proceeding progress in wood and bio-fibre

plastic composites international conference, Toronto

4. Stark NM, Matuana LM (2004) Polym Degrad Stab 86:1–9

5. Stark NM, Matuana LM, Clemons CM (2004) J Appl Polym Sci

93:1021–1030

6. Stark NM, Matuana LM (2006) Polym Degrad Stab 91:3048–

3056

7. Muasher M, Sain M (2006) Polym Degrad Stab 91:1156–1165

8. Stark NM, Matuana LM (2007) Polym Degrad Stab 92:1883–

1890

9. Fabiyi JS, McDonald AG, Wolcott MP (2005) In: Proceeding of

the 8th international conference on woodfiber-plastic composites,

Madison, WI, pp 191–196, May 23–25

10. Sears KD, Jacobson R, Caulfield DF, Underwood J (2001) In:

Proceedings of the 6th International Confrence on Wood-plastic

composites, Madison, WI, pp 27–34

11. Philip M, Attwood J, Hulme A, Williams G, Shipton P (2004)

The waste and resources action programme, p 113

12. Torik AA, Shirakawa H, Nagaya S (1990) J Appl Polym Sci

40:1637–1646

13. Isasi JR, Mandelkern L, Galante MAJ, Alamo RG (1999) J Polym

Sci Part B: Polym Phys 37:323–334

14. Technical Association of Pulp and Paper Industry (TAPPI T 204

om-88) (1996) Atlanta, Georgia, pp 1–3

15. Wise LE, Murphy M, D’ Addireco AA (1946) Paper Trade J

122:35–43

16. American Society Testing and Materials (ASTM D 790-00)

(2001) Annual book of ASTM standards. Conshohocken, PA,

08(01):145–153

17. American Society Testing and Materials (ASTM D 6662) (2003)

Annual book of ASTM standards. Conshohocken, PA,

08(03):1035–1049


ASTM annual book. Conshohocken, PA, 08(03):308–312


Annual book of ASTM standards. Conshohocken, PA, 06(01):

240–249

20. CIE (1986) CIE publication number 15.2, 2nd edn. Vienna,

Austria, p 74

21. Tidjani A (2000) Polym Degrad Stab 68:465–469

22. Mayo DW (2003) In: Mayo DW, Miller FA, Hannah RW (eds)

Course Notes on the interpretation of infrared and Raman spectra.

John, USA, pp 179–204

23. Lacoste J, Carlsson DJ (1992) J Polym Sci Part A Polym Chem

30:493–500

24. Lacoste J, Carlsson DJ, Falicki S, Wiles DM (1991) Polym

Degrad Stab 34:309–329

25. Fabiyi JS, McDonald AG, Wolcott MP, Griffiths PJ (2008) Polym

Degrad Stab (in press)

26. Kamdem DP, Rield B, Adnot A, Kaliaguine S (1991) J Appl

Polym Sci 43:1901–1912

27. Matuana LM, Kamdem DP (2002) Polym Eng Sci 42:1657–1666

28. McDonald AG, Clare AB, Dawson D (1999) In: Proceeding

of the 53rd appita annual confrence, Rotorua, New Zealand 1,

pp 51–56

29. Schauwecker C, Morrell JJ, McDonald AG, Fabiyi JS (2006) For

Prod J 56:123–129


Philadelphia, PA, 8(03):337–341

31. Kodjie SL, Li L, Li B, Cai W, Li CY, Keating M (2006) J

Macromol Sci-Part B: Phys 45:231–245

32. Supaphol P, Spruiell JE (2002) J Appl Polym Sci 86:1009–1022

33. Selden R, Nystrom B, Langstrom R (2004) Polym Comp 25:543–

553

34. Rabello MS, White JR (1997) Polym 38:6379–6387

35. Stark NC (2003) PhD thesis, School of Forest Research and

Environment Science, Michigan Technical University, p 217

36. Faix O (1992) In: Lin SY, Dence CW (eds) Methods in Lignin

Chemistry. Springer-Verlag, Berlin, pp 83–132

37. Mononen K, Jaaskelainen AS (2005) Holzforschung 59:381–388

38. Dorris GM, Gray DG (1978) Cellul Chem Tech 12:721–734

39. Grigsby W, McDonald AG, Thumm A, Loxton C (2004) Holz

Roh ahs und Werkstoff 62:358–364

40. Freudenberg K, Neish AC (1968) Constitution and biosynthesis

of lignin. Springer Verlag, Berlin

41. Lojewska J, Miskowiec P, Lojewski T, Proniewicz LM (2005)

Polym Degrad Stab 88:512–520

42. Meier D, Faix O (1992) In: Lin SY, Dence CW (eds) Springer-

Verlag, Berlin Heidelberg, pp 177–199

43. Robertson JE (2001) PhD Dissertation, Virginia Polytechnic

Institute and State University, p 185

44. Chiantore O, Tripodi S, Sarmoria C, Valle AE (2001) Polym

42:3981–3987

45. Hoekstra HD, Spoormaker JL, Breent J, Audouin L, Verdu J

(1995) Polym Degrad Stab 49:251–262

46. MacCallum JR (1965) Makromol Chem 83:129–136

47. Jabarin SA, Lofgren EA (1994) J Appl Polym Sci 53:411–423

48 J Polym Environ (2009) 17:34–48

123

Wood Modification Effects on Weathering of HDPE-Based Wood Plastic Composites

Documents