Composites: Part B 50 (2013) 259–264

Contents lists available at SciVerse ScienceDirect

Composites: Part B
journal homepage: www.elsevier.com/locate/compositesb

Properties of particleboard made from rubberwood using modified starch as binder

Mohd Hazim Mohamad Amini a, Rokiah Hashim a,⇑, Salim Hiziroglu b, Nurul Syuhada Sulaiman a,
Othman Sulaiman a
a
Division of Bio-resource, Paper and Coatings Technology, School of Industrial Technology, Universiti Sains Malaysia, 11800 Penang, Malaysia
b
Department of Natural Resource Ecology & Management, 303G Agricultural Hall, Okhahoma State University, Stillwater, OK 74078-6013, USA

a r t i c l e i n f o a b s t r a c t

Article history: The objective of the study was to evaluate physical and mechanical properties of experimental
Received 11 April 2012 particleboard panels made from rubberwood (Hevea brasiliensis) using modified starch as binder. Panels
Received in revised form 8 November 2012 were manufactured using 15% corn starch modified with glutardialdehyde and tested for their properties
Accepted 24 February 2013
based on Japanese Standard. The modulus of rupture and the internal bond strength of the panels met the
Available online 4 March 2013
requirement of the specified standard. Based on the findings in this work modified corn starch can have a
potential to be used as binder to produce particleboard panels with acceptable properties.
Keywords:
Ó 2013 Elsevier Ltd. All rights reserved.
A. Wood
D. Mechanical testing
D. Surface analysis

1. Introduction formaldehyde which could be easily released by hydrolytic

reactions [3]. Formaldehyde emission from urea formaldehyde
Rubberwood (Hevea brasiliensis) is one of the most commonly bonded panels gained attention as a public health concern last
used raw materials to manufacture composite panels such as fibre- 30 years. It is well known fact that formaldehyde causes a signifi-
board and particleboard in Malaysia. Rubberwood originated as cant health problems as well as environmental pollution. Of course
indigeneous species to the Amazon forest in Brazil was first intro- very low concentration of formaldehyde in the atmosphere does
duced to South East Asia in mid 1800s [1]. In early 1990s Malaysia not create any problems. For example, typical formaldehyde con-
had a great success using rubberwood in manufacture of value- centration in atmosphere is generally less than 0.1 ppm [4]. In
added products and became leader in South East Asia. The rubber- one of the preview works, free formaldehyde percentage in urea
wood sawn timber industry in Malaysia is well developed and used formaldehyde bonded particleboard made from different European
such resource as efficiently as possible. Currently waste materials species were found less than 0.3 ppm. Also experimental particle-
from furniture and lumber manufacture and low quality small logs boards panels manufactured from pine and spruce resulted in low-
are the main raw material for composite panel producers in er formaldehyde emission than those samples made from beech
Malaysia. [5]. Wood composite industries in many countries try to control
Similar to wood composite industry in many other countries, and reduce formaldehyde emission from the wood composite pan-
formaldehyde based adhesives are also widely used in Malaysia. els. There maybe two approaches to achieve that, namely to modify
In the case of particleboard, urea formaldehyde is the most com- the chemical structure of the adhesive and reduce the amount of
monly used binder due to its fast curing time, clear color and resin in board manufacture. However, even very little reduction
low cost [2]. However one important disadvantage of such adhe- of adhesive in the panels can significantly influence both physical
sive is its formaldehyde emission. The fundamental mechanism and mechanical properties of the final product. Therefore some
in formaldehyde emission from urea formaldehyde bonded parti- manufacturers are also interested in development and using non-
cleboard is simply related to unreached free formaldehyde from formaldehyde base adhesive in their product line to eliminate such
the binder and hydrolysis of partially and completely cured adhe- problem. Green and environmentally friendly materials including
sive. Several conditions of formaldehyde could be present such as soybean and various types of starches would have potential to
monomeric formaldehyde entrapped between wood particles, as produce composite panels without having problems stated above.
monomeric by hydrogen bonding of formaldehyde to the wood Starch is carbohydrate materials that consist of amylase and
or as polymeric (solid) formaldehyde as well as loosely bound amylopectin which could be differentiated by its chemical struc-
ture. The linear a-(1 ? 4) linked glucan is called amylase while
⇑ Corresponding author. Fax: +60 46573678. an a-(1 ? 4) linked glucan with 4.2–5.9% a-(1 ? 6) branch link-
E-mail address: hrokiah@usm.my (R. Hashim).
ages is amylopectin [6]. It can be obtained from various plant

1359-8368/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.compositesb.2013.02.020
260 M.H.M. Amini et al. / Composites: Part B 50 (2013) 259–264

materials such as corn, potato, rice, wheat, sago and many more on 15 samples with 3 readings taken for each samples using the Hom-
and widely available throughout the world commonly used in food mel Tester T500, which consists of the main unit and the pick up model
industries. Therefore, modification of starches were well docu- TkE. The pick up has a skid type diamond stylus with 5 lm tip radius
mented by many researchers [7]. Various modifications of starch and a 90° tip angle. The stylus moves over the surface at constant speed
were evaluated, including through oxidation, esterification, ether- of 1 mm/s over 15.2 mm tracing length. Vertical displacement of the
ification and crosslinking of starch. These processes yield, for stylus was converted into an electrical signal. Three roughness param-
example carboxymethyl starch, dialdehyde starch, hydroyethyl eters, i.e., average roughness (Ra), mean peak-to-valley height (Rz), and
starch and starch xanthate [8]. Besides for food industries applica- maximum roughness (Rmax) were used for surface roughness evalua-
tion, utilization of starch in non-food industries was also investi- tion of the samples. Specifications of these parameters were discussed
gated by researchers, especially in the biodegradable in previous studies [5,13,14]. Determination of surface roughness is
thermoplastic field [9]. Although properties of rubberwood parti- important as some coating materials really depend on the surface of
cleboard have been investigated in various works, currently there the panels to work. Scanning electron microscopy (SEM) analysis
is no information on characteristics of composite panels made was also carried out to see the interaction and distribution of the adhe-
from rubberwood manufatured by modified starch as binder sive between the wood particles.
[1,10]. Therefore the objective of the work was to manufacture Characterizations of chemical and thermal properties of the
experimental panels from rubberwood and modified starch, and manufactured particleboard were done. Infra-red spectra in the
to evaluate both physical and mechanical properties of the samples range of 4000–470 cm1 of the particleboard were measured with
to determine if they were similar to those of commercially manu- FTIR spectrophotometer, (Nicolet, AVATAR FTIR-360), running Om-
factured particleboard panels. nic software, to characterize the functional group inside the parti-
cleboard. To evaluate the crytallinity of the materials inside the
particleboard, finely powdered samples were prepared from the
2. Materials and methods IB test specimen, examined by XRD analysis using Diffraktometer
D5000 Kristalloflex, Siemens. Step scan measurements were done
Commercially produced rubberwood ( H. brasiliensis) particles using X-rays (Cu–Ka) at 40 kV and 40 mA. Scanning of 2h was ran-
supplied by a local particleboard company in Negeri Sembilan, ging from 10.0° to 40.0° corresponding to scanning speeds of 0.02°/
Malaysia were used to make experimental panels. The particles min and 2°/min [15]. The crystallinity index (C Ir) was calculated
were dried to 2% moisture content in a laboratory type oven. Corn using the formula:
starch in powder form modified with glutaraldehyde in liquid form
in a ratio of 1:2 (w/w) was used as binder in a ratio of 15% based C Irð%Þ ¼ ðI200  Iam Þ=I200  100 ð1Þ
oven dry particle weight. Initially corn starch powder was dis-
solved in distilled water with a temperature of 30 °C before it K k
L¼ ð2Þ
was stirred and added 25% glutardialdehyde solution. Glutardial- b  cos h
dehyde (GDA) is a colorless oily liquid organic compound with
where I200 is the peak intensity corresponding to crystalline and
the formula of CH2 (CH2CHO)2. It is widely used as disinfector
Iam is the peak intensity of the amorphous fraction.
agent for medical equipment. It has specific density ranging from
Thermal decomposition of the manufactured particleboards
1.06 to 1.12 at 20 °C [11].
was done using a Metler Toledo TGA/SDTA 851 thermogravimetric
A total of 15 panels, five for each density level with dimension
analyzer, recorded from 30 °C to 800 °C for samples of 5–10 mg
of 20.1 cm by 20. 1 cm by 0.5 cm was manufactured for the exper-
placed in an aluminum pan with a heating rate of 20 °C min1 un-
iments, as shown in Table 1. Panels were made for target density
der a nitrogen atmosphere [16]. Differential scanning calorimetry,
levels of 0.60 g/cm3, 0.70 g/cm3 and 0.80 g/cm3. Fifteen percent
(DSC) Pyris 1, Perkin Elmer was used to evaluate the thermal
modified corn starch was manually mixed with rubberwood parti-
behaviors of the manufactured particleboards, with heating rate
cles before they were processed in a computer control press using a
of 10 °C/min. About 7 mg of powdered particleboard was added
pressure of 5 MPa at a temperature of 165 °C for 20 min. Panels
to an aluminum pan and sealed. An empty, sealed aluminum pan
were conditioned in a climate chamber with a temperature of
was used as reference. Then it was heated from 10 °C to 280 °C
20 °C and a relative humidity of 65% for 2 weeks. After the samples
at the respective heating rate. Melting temperature of particle-
were conditioned, their modulus of elasticity (MOE), modulus of
board was determined from the obtained DSC curve.
rupture (MOR), internal bond strength (IB), thickness swelling,
water absorption and surface roughness were evaluated based on
Japanese Industrial Standard [12]. 3. Results and discussion
Both bending and internal bond strength test were carried out on
using Tensile Strength Tester Machine Model 5582 (INSTRON) Table 2 displays test results of the samples. The highest average
equipped with a load cell having 1000 kg each. Crosshead speeds of values of 3541 N/mm2 and 20.38 N/mm2 were found for MOE and
10 mm/min and 2 mm/min were used for bending strength and inter- MOR of the panels with density of 0.80 g/cm3, respectively. These
nal bond strength, respectively. Thickness swelling and water absorp- values were 59.41% and 58.35% higher than those of the specimens
tion test were carried out by soaking 50 mm  50 mm  5 mm made with 0.70 g/cm3 target density. The samples produced with
samples in water for 2 h and 24 h. Surface roughness profile were done 0.60 g/cm3 density level resulted in the lowest bending properties

Table 1
Experimental design.

Particleboard density (g/cm3) Number of sample

Density Thickness swelling Water absorption Surface roughness Bending Internal bond strength
0.60 30 15 15 15 15 15
0.70 30 15 15 15 15 15
0.80 30 15 15 15 15 15
 M.H.M. Amini et al. / Composites: Part B 50 (2013) 259–264 261

Table 2
Test results with statistical analysis of the samples.

Target density Measured Bending test (N/mm2) Internal bonding Thickness swelling, % Water absorption, % Surface roughness, lm
(g/cm3) density (g/cm3) (N/mm2)
MOE MOR 2h 24 h 2h 24 h Ra Rz Rmax
0.60 0.58 (0.03)a 1967.07 9.85 0.62 (0.11)a 28.76 39.087 87.35 106.01 3.62 41.69 50.65
(349.50)a (2.43)a (6.75)a (5.92)a (8.55)a (3.9)a (0.97)a (8.28)a (13.15)a
0.70 0.69 (0.04)b 2221.28 12.87 0.88 (0.21)b 34.81 47.53 90.02 107.42 3.64 37.51 52.92
(336.44)a (2.44)b (7.38)b (10.64)b (11.26)a (8.31)a (1.23)a (8.63)ab (11.1)a
0.80 0.78 (0.07)c 3540.97 20.38 1.02 (0.26)b 45.77 60.80 91.09 107.58 2.86 35.15 50.3
(475.22)b (3.55)c (7.32)b (8.13)b (6.37)a (6.89)a (1.13)b (9.13)b (9.07)a

Values in parentheses shows standard deviation.

Different letter shows significant difference between groups within the same column at a = 0.05.

between particles during pressing and cured as function of temper-

ature. Fig. 1 illustrates modified corn starch granules located on the
particles. Statistical analysis shown in Table 2 indicated that IB
strength values of panels from 0.60 g/cm3 target density were sig-
nificantly different from the panels made from other two densities
group. This analysis showed the maximum IB strength could be ob-
tained in a range of densities between 0.70 g/cm3 and 0.80 g/cm3
where higher density of panel could only slightly increased the
IB strength.
Starch is hydrophobic material and even if it is modified by
using GDA, it still keeps such characteristic, influencing dimen-
sional stability of the panels adversely. Both thickness swelling
and water absorption of the panels were found unsatisfactorily
and did not meet minimum requirement stated in the JIS standard.
Thickness swelling of the samples ranged from 28.76% to 60.80% as
a result of 2 and 24 h water soaking test. Substantial amount of
water also ranged from 87.35% to 107.58% for the same exposure
time periods. Results in Table 2 showed poor dimensional stability
Fig. 1. SEM cross sectional view of particleboard made using glutardialdehyde of the panels made in this work. For water absorption, all the den-
modified corn starch as a binder at 500 magnification with the presence of
sity groups were not significantly different from each other. This
modified corn starch granules shown with arrows.
could be explained because panels from each density absorbed al-
most same amount of water, in ratio to their original sample
weight. The thickness swelling follows the same trend as IB
having average values of 1967 N/mm2 and 9.85 N/mm2 for MOE strength of panels with 0.60 g/cm3 target density was significantly
and MOR tests. It is well known fact that density of almost any different from the other two densities group.
wood composites is a major parameter influencing their bending Dimensional stability of the experimental particleboards and fi-
characteristics [17]. All density levels of the panels made 15% bers is a main concern when the panels were made with starch or
was modified starch met JIS standard in term of their bending without any binders [19]. Such problem became more prominent
properties which prescribed to be at least 8.0 N/mm2. In a previous in the case of starch used due to its hygroscopicity as mentioned
study, particleboard made from oil palm trunk particles without above. However, certain treatments including heat treatment and
using any binders had 13.37 N/mm2 for their MOR [18]. It seems chemical treatments could be considered to enhance thickness
that using starch as binders enhanced overall bending properties swelling and water absorption of the samples. Also in typical com-
of the panels as compared to those made without using any adhe- mercial panel manufacturer, around 1% wax is used to have better
sive. Effect of density was also main factor on IB strength values of dimensional properties of the products. In this study, if very little
the samples. The panels made with 0.80 g/cm3 density had the amount of wax was added into the samples, particularly their poor
highest IB strength value of 1.02 N/mm2 while the lowest IB value thickness swelling and water absorption could have been improved.
of 0.62 N/mm2 was determined from those manufactured with Panels with density levels of 0.60 g/cm3 and 0.70 g/cm3 had
0.60 g/cm3 target density. Table 2 shows statistical analysis of all average roughness values of 3.62 lm and 3.64 lm, respectively.
tests. The MOE values from board made of 0.60 g/cm3 target den- Based on statistical analysis, no significant difference was found
sity were determined was not significantly different from board between surface roughness values of these two types of samples.
made with target density of 0.70 g/cm3 panels. However, the However, once their density was increased to 0.80 g/cm3 surface
MOE value of panels made at 0.80 g/cm3 target density were signif- quality of the panels significantly improved, having 2.86 lm aver-
icantly different from MOE values from panels made with target age value, 26.57% lower than those panels made with 0.60 g/cm3
density of 0.60 g/cm3 and 0.70 g/cm3. For MOR, panels from each density. It is well known fact that surface quality of particleboard
density group were significantly different from each other. This increases with its increasing density due to higher densification
observation support the theory of density increment will increase and compaction ratios of the face layer. In another study, particle-
the bending properties of wood panels. board made with higher density also resulted in better surface
Satisfactory bending and IB strength properties of the panels quality [20].
suggested that starch was mixed uniformly with particle and Characterization of grounded particleboard powder using FT-IR
developed a well bonding line. Based on micrograph taken from shows presence of O–H group at wavenumber of 2928.69 cm1,
the surface of samples it is clear that starch particles were attached that was detected from wood and starch structure. Carbonyl chro-
to the surface of particles and resulted in a complete contact mophores were detected at wavelength 1736.41 cm1 and
262 M.H.M. Amini et al. / Composites: Part B 50 (2013) 259–264

1736.41

1646.46

2928.69

B
I002

Secondary
Secondary
peak
peak

Iam

Fig. 2. FT-IR spectra (A) and XRD spectra (B) of glutardialdehyde modified starch particleboard sample made using glutardialdehyde modified corn starch as a binder.

1646.46 cm1, in aldehyde and ester form, respectively, as shown sharp decrease of weight as the sample was heated from 250 °C
in Fig. 2A. These chromophores shows the presence of glutardial- to 425 °C which is from 91.7% to 25.69%, a 66.01% weight decre-
dehyde modified corn starch as the binder in the particleboard ment. This large weight loss was probably due to the thermal
[21]. An XRD spectrum in Fig. 2B. shows that the powder of glutar- decomposition of hemicellulose, carbon dioxide and water [16].
dialdehyde modified corn starch was highly amorphous. Major The TG profile also showed that the ash content of the particle-
intensity peaks at 2h was observed near 23° related to their crystal- board is lower than 18.43% as longer heating and higher tempera-
line structure. Secondary peaks were also observed at 2h = 15° and ture could increase decomposition of sample. The DTG profile
2h = 20.5°. The crystallinity index was calculated to quantify the showed a detail view of the analysis. The DTG shows rate of
crystallinity of the sample, yielded as 70.59%. decomposition or mass loss at certain temperature [22]. Water loss
Fig. 3A and B illustrate the weight loss curves (TG) and deriva- is highest at 43 °C while thermal degradation of sample was at the
tive thermogravimetric (DTG) profiles of the modified corn starch highest rate at 360 °C, determined as 0.15 %/min. From differential
particleboard. A slight 6% weight loss between the 30 °C and scanning calorimetry analysis of manufactured particleboard sam-
100 °C was due to loss of moisture from the sample. There was a ples, two exothermic peaks were found at 32 °C and 172 °C. There
 M.H.M. Amini et al. / Composites: Part B 50 (2013) 259–264 263

A 4. Conclusions
100 91.7%
From the results, the highest and lowest modulus of elasticity
values were determined as 3540 MPa and 1967 MPa for the panels
80 with 0.80 g/cm3 and 0.60 g/cm3 density levels, respectively. Inter-
nal bonding strength showed the same trend, highest at 1.02 N/
Weight (%)

mm2 for 0.80 g/cm3 panels and lowest at 0.62 N/mm2, measured
60
from 0.6 g/cm3 panels. Thus, it can be concluded that these panels
satisfied the Japanese Industrial Standard. The dimensional stabil-
40 ity of the panels towards moisture needs to be improved by
addition of water repellent materials or the manufactured particle-
25.69%
boards should be restricted to only dry condition usage. Consider-
20 ing all the parameters above, glutardialdehyde modified corn
starch has the possibility to be commercialized as wood binder
50 100 150 200 250 300 350 400 450 500 550 600 650 700 in particleboard industry.
Temperature (°C)

B Acknowledgement
0.00
The authors were thankful for the Ministry of Higher Education
of Malaysia for the MyPhd funding to Mohd Hazim Mohamad Ami-
Derivative weight (1/min)

ni. The authors also acknowledged Universiti Sains Malaysia for the
-0.05 Research University Grant (1001/PTEKIND/815066) to Rokiah Ha-
shim and Postgraduate Research Grant Scheme (1001/PTEKIND/
844104) for Mohd Hazim Mohamad Amini. Appreciation was also
-0.10 given to the Heveaboard Sdn Bhd for providing the raw materials
for particleboard making.

-0.15 References

[1] Teoh Y, Don M, Ujang S. Assessment of the properties, utilization, and

50 100 150 200 250 300 350 400 450 500 550 600 650 700 preservation of rubberwood (Hevea brasiliensis): a case study in Malaysia. J
Temperature (°C) Wood Sci 2011;57(4):255–66.
[2] Pizzi A. Urea-formaldehyde adhesives. In: Pizzi A, Mittal KL, editors. Handbook
of adhesive technology. New York: Taylor & Francis Group, LLC; 2003 [Revised
Fig. 3. TG curve (A) and DTG curve (B) of the manufactured particleboard sample
and expanded].
made using glutardialdehyde modified corn starch as a binder.
[3] Que Z, Furuno T, Katoh S, Nishino Y. Effects of urea-formaldehyde resin mole
ratio on the properties of particleboard. Build Environ 2007;42(3):1257–63.
[4] Markessini E. Formaldehyde emission from wood-based panels and ways to
reduce them. Monument Environ 2004;2:57–64.
[5] Akbulut T, Hiziroglu S, Ayrilmis N. Surface absorption, surface roughness, and
10 formaldehyde emission of Turkish MDF. Forest Prod J 2000;50(6):45–8.
[6] Robyt JF. Starch: structure, properties, chemistry and enzymology. In: Fraser-
Reid OB, Tatsuta K, Thiem J, editors. Glycoscience chemistry and chemical
biology. Berlin (Heidelberg, New York): Springer-Verlag; 2008. p. 2866.
8 [7] Verwimp T, Vandeputte GE, Marrant K, Delcour JA. Isolation and
Heat Flow Endo Up (mW)

characterisation of rye starch. J Cereal Sci 2004;39(1):85–90.

[8] Tharanathan RN. Starch—value addition by modification. Crit Rev Food Sci Nutr
2005;45(5):371–84.
[9] Thiré RMSM. Starch-based plastics. In: Bertolini AC, editor. Starches :
6
characterization, properties, and applications. Boca Raton, FL: CRC; 2010.
[10] Hong LT, Sim HC, Institute FR, Rubberwood processing and utilisation1999:
Forest Research Institute of Malaysia.
[11] Lenga RE. The sigma-aldrich library of chemical safety data. vol. 2. United
4 Kingdom: Sigma-Aldrich Corporation; 1988. p. 2175.
[12] JIS A 5908. Particleboards, Japanese standards association: Tokyo, Japan; 2003.
[13] American National Standards Institute. Surface texture: surface roughness,
waviness, and lay : ASME B46.1-1995 (revision of ANSI/ASME B46.1-1985).
2 New York. American Society of Mechanical Engineers; 1995. p. 43.
[14] American National Standards Institute. Medium density fiberboard (MDF) for
interior applications. Premiere Ct. Gaithersburg, MD: Composite Panel
0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 Association; 2009.
Temperature (°C) [15] Hermawan D, Hata T, Kawai S, Nagadomi W, Kuroki Y. Manufacturing oil palm
fronds cement-bonded board cured by gaseous or supercritical carbon dioxide.
J Wood Sci 2002;48(1):20–4.
Fig. 4. DSC curve of glutardialdehyde modified corn starch particleboard sample
[16] Baskaran M, Hashim R, Said N, Raffi SM, Balakrishnan K, Sudesh K, et al.
made using glutardialdehyde modified corn starch as a binder.
Properties of binderless particleboard from oil palm trunk with addition of
polyhydroxyalkanoates. Compos Part B: Eng 2011.
[17] Wong ED, Zhang M, Wang Q, Kawai S. Formation of the density profile and its
effects on the properties of particleboard. Wood Sci Technol 1999;33(4):
is no specific explanation on these peaks but loss of moisture or
327–40.
reaction of unreacted glutardialdehyde or chemical compound [18] Hashim R, Nadhari WNAW, Sulaiman O, Kawamura F, Hiziroglu S, Sato M, et al.
could be the reason. Highest endothermic peak was found as Characterization of raw materials and manufactured binderless particleboard
8.63 mW, at temperature of 92 °C, which is the melting tempera- from oil palm biomass. Mater Des 2011;32(1):246–54.
[19] Ayoub AS, Rizvi SSH. An overview on the technology of cross-linking of starch
ture of the particleboard, as shown in Fig. 4. for nonfood applications. J Plast Film Sheet 2009;25(1):25–45.
264 M.H.M. Amini et al. / Composites: Part B 50 (2013) 259–264

[20] Hiziroglu S, Kosonkorn P. Evaluation of surface roughness of Thai medium [22] Hermawan D, Hata T, Umemura K, Kawai S, Nagadomi W, Kuroki Y. Rapid
density fiberboard (MDF). Build Environ 2006;41(4):527–33. production of high-strength cement-bonded particleboard using gaseous or
[21] Wade LG. Organic chemistry. 6th ed. New Jersey: Pearson Education Inc.; 2006. supercritical carbon dioxide. J Wood Sci 2001;47(4):294–300.
p. 1330.