Medium density fiberboard (MDF) is one of the extensively used wood
based materials as part of building and housing furniture. MDF however
is prone to fire hazard and an enhanced resistance to fire is therefore
desired (Hashim et al., 2005; Chih and Szu, 2003; Laufenberg et
al., 2006; Barnes and Farrell, 1978). One of the major considerations
in the manufacture of flame retardant MDF is maintaining the necessary
physical and mechanical properties. Factors such as wood species, moisture
content, pressing conditions, flame retardant treatment critically affect
these properties of the panels (Ayrilmis et al., 2007; Woo and
Schniewind, 1987; Berndt et al., 1990).
Zinc borate has been used as flame retardant for wood and wood products
(Kozlowski et al., 1999; LeVan and Winandy, 1990; Garba, 1999).
Aluminum trihydrate is widely used as flame retardant additives for plastics
and elastomers (Laufenberg et al., 2006; Sain et al., 2004).
Work on the use of hydrated alumina in MDF has been carried out by Barnes
and Farrell (1978) indicating its potential. The study was conducted with
5 and 10% hydrated alumina and 8% urea formaldehyde based on dry fiber.
Sodium aluminate is an important industrial inorganic chemical. It is
used in water treatment and as a source of aluminum in synthetic applications.
It is often used as the aluminum source in the preparation of zeolites
and other catalytic materials (Misra, 1986). Aluminum trihydrate and sodium
aluminate are not usually used as a flame retardant system for wood products.
In this study we investigated the effect of incorporation of these flame
retardants on the properties of MDF made using phenol as adhesive.
MATERIALS AND METHODS
Experimental medium density fiberboard (MDF) 21.2x21.2 cm by 0.5
cm with an average density of 0.7 g cm-3 were made using a
small scale laboratory press in 2005. The boards were made from thermo
mechanical processed rubberwood (Hevea brasiliensis) fibers from
an MDF mill in Malaysia. Phenol formaldehyde (PF) resin obtained from
Hexion Specialty Chemicals was used at 15% resin level based on oven dry
weight of the boards. The solid content of adhesive was 56% and its viscosity
was 85 mPas cps-1 at 25Â°C with a pH 13.
Experimental medium density fiberboard (MDF) 21.2x21.2 cm by 0.5 cm with
an average density of 0.7 g cm-3 were made using a small scale
laboratory press. The boards were made from thermo mechanical processed
rubberwood (Hevea brasiliensis) fibers from an MDF mill in Malaysia.
Phenol formaldehyde (PF) resin obtained from Hexion Specialty Chemicals
was used at 15% resin level based on oven dry weight of the boards. The
solid content of adhesive was 56% and its viscosity was 85 mPas cps-1
at 25Â°C with a pH 13.
Thickness swelling and water absorption of MDF were done in accordance
to ISO/DIS 16983 (2002). Modulus of rupture (MOR) or bending strength
was done in accordance to International Standard ISO/DIS 16978 (2002)
with size of 12.0x2.0x0.5 cm. The internal bond strength (IB) was evaluated
in accordance to International Standard ISO/DIS 16984 (2002).
In order to understand more on the effect of flame retardant on the resin
used, a separate study on resin and flame retardant mix were done. Phenol
formaldehyde resin was mixed with flame retardant chemical at various
flame retardant concentrations (10, 15, 20 and 30%). The adhesive mix
was then made into thin layer films. The films produced were then evaluated
A modified Cabinet Method Test (ASTM D 1360-79, 1979) was used to evaluate
relative flame retardant properties using samples of size 21.2x21.2x0.5
cm. This test included char index, weight loss and area of ellipse. Thermal
weight loss measurements were made using thermo gravimetric analyzer (TGA)
(Perkin-Elmer TGA 7). Testing was carried out under a stream of dry N2
gas/O2 gas with a flow rate of 30 mL min-1 at a
temperature ranging of 30-800Â°C with a heating rate of 20Â°C min-1.
Differential scanning calorimeter (DSC) analysis was carried out using
Perkin-Elmer Pyris-1 DSC, equipped with an internal Cooler 2P-cooling
accessory and calibrated using n-decane and indium. Samples of 5 mg each
were examined within an atmosphere of dry N2 gas maintained
at a flow rate of 20 mL min-1. Samples were encapsulated in
standard aluminum pans and an empty aluminum pan used as a reference.
All samples were annealed at a heating rate of 20Â°C min-1.
The location and presence of flame retardant chemical in fiber were observed
using Leo Supra 55 Vp Ultra-high resolution analytical Field Emission
Scanning Electron Microscopy (FESEM) and EDAX using split samples.
RESULTS AND DISCUSSION
Physical Properties: The presence of flame retardant chemicals was indicated based on
the EDAX analysis showing various flame retardant chemicals. The flame
retardant seems to be distributed well in the board (Fig.
Moisture content for the panels ranged from 7 to 11%, while the mean
value for specific gravity range from 0.74-0.77 (Table 1).
For thickness swell after 24 h, there was a progressive decrease in swell
||SEM micrographs and EDAX spectrum of flame retardant MDF (a)
sodium aluminate, (b) zinc borate and (c) aluminum trihydrate
for all types of flame retardant chemicals as the percentage concentration
of the flame retardant chemicals increased. Flame retardant MDF incorporated
with 30% zinc borate had the lowest thickness swell. For thickness swell
after cyclic, the swell increased progressively as the amount of flame
retardant chemical increased.
||Physical and mechanical properties of flame retardant MDF
|aNo. in parentheses are standard deviation,
FR: Fire retardant, MC: Moisture content, SG: Specific gravity, TS:
Thickness swelling after 24 h soaking, WA: Water absorption, IB: Internal
bond, MOR: Modulus of rupture, Ctrl: Control, SA: Sodium aluminate,
ZB: Zinc borate, ATH: Aluminum trihydrate
Flame retardant MDF incorporated with 10% aluminum trihydrate indicated
the lowest mean thickness swell values after a cyclic test, whereas boards
incorporated with 20% zinc borate, 30% zinc borate and 30% sodium aluminate
showed high mean thickness swell after a cyclic test.
The results for water absorption correspond well with the thickness swell
where as the amount of flame retardant chemicals increased, water absorption
and thickness swell reduced. Panels incorporated with 30% aluminum trihydrate
showed the least water absorption. Flame retardant MDF incorporated with
10% zinc borate showed the lowest mean of water absorption after a cyclic
test. High mean water absorptions after a cyclic test were seen in boards
made using 20 and 30% zinc borate and 30% aluminum trihydrate. Some flame
retardant chemicals probably penetrated the rubberwood fibers during the
production of MDF which would hinder water absorption by the samples (Fig.
1). In general, the higher the level of flame retardant chemical the
lower is the water absorption.
Mechanical Properties: In general modulus of rupture (MOR) increased significantly at p
= 0.05 as the concentration of the flame retardant chemicals increased
with boards treated with 30% aluminum trihydrate had the highest MOR (Table
1). The higher MOR for higher percentage of flame retardant probably
resulted from the formation of crystals in the boards might improved the
MOR. The presence and location of the flame retardant chemicals in the
treated boards are shown in Fig. 1. Similar minimal improvement
was also reported by Hashim et al. (2005). However, when the treated
boards were subjected to a cyclic test, the MOR progressively decreased
The internal bond strength (IB) of flame retardant MDF is shown in Table
1. Results after underwent cyclic conditions are also shown in Table
1. There was no general trend in the effect of the
||Cabinet test and thermogravimetry (TGA) analysis of flame
|aNo. in parentheses are standard deviations;
SA: Sodium aluminate; ZB: Zinc borate, ATH: Aluminum trihydrate
treatment in these tests when the boards were treated with 10% flame
retardant level. As the concentration of the chemicals increased all the
IB values progressively decreased. The reason for this probably there
is a disturbance in the adhesion of the board with the presence of flame
retardant chemicals. In order to understand more on this, a separate study
was done on the bonding characteristics using the cast film tests. Films
made from a mixture of phenol formaldehyde and zinc borate indicated visible
signs of phase separation and cluster formation, leading to discontinuity
in the film. The difference in film formation may explain why MDF incorporated
with aluminum trihydrate and sodium aluminate using phenol formaldehyde
resin had better IB after cyclic and boil than flame retardant MDF incorporated
with zinc borate.
Flame Retardant and Thermal Properties: There is significant effects
of the flame retardant performance on the type of flame retardant chemicals
as can be seen in Cabinet test and TGA analysis (Table 2).
When board exposed to heat, it undergoes pyrolysis and chars and this
produce tar and combustible gases. However, when board are treated with
flame retardant chemical, it also undergoes pyrolysis and chars but very
minimal tar and combustibles gas released (Abdul Rashid and Murphy, 1993).
Generally, the char index decreased as the levels of flame retardant
chemical increased. Flame retardant MDF incorporated with zinc borate
has the highest value of a char index, weight loss and area of ellipse.
It showed that zinc borate was less efficient in reducing flame propagation
in comparison to sodium aluminate and aluminum trihydrate.
The weight loss of sample during the cabinet test indicated that boards
treated with sodium aluminate showed the best performance with the least
weight loss even though the other two flame retardants also imparted good
The results showed that as the loading of flame retardant increased,
the area of ellipse decreased. Boards incorporated with zinc borate had
the largest area of ellipse of 98.31 cm2 followed by boards
treated with aluminum trihydrate and sodium aluminate with value of 80
and 52.62 cm2 for 10% flame retardant levels, respectively.
Control samples showed considerable weight loss of 73.7% compared to
samples incorporated with flame retardants. It showed that boards treated
with sodium aluminate had the lowest mean weight loss even at 10% concentration
level (58.5%). This indicated the efficacy of the flame retardant chemical.
The percentage weight loss of MDF was influenced more by the type of flame
retardants than the level of flame retardant.
||TGA curves for Flame retardant MDF made from rubberwood
(RW) incorporated with flame retardant chemicals using phenol formaldehyde
resin (a) At 10% flame retardant concentration (b) At 30% flame retardant
concentration, SA: Sodium aluminate, ZB: Zinc borate, ATH: Aluminum
The percentage weight loss and char residue (%) represent the amount
of carbon that are characteristic of boards incorporated with flame retardant
chemicals. High char residues are usually associated with stable thermal
structures in the backbone of the board.
A remarkable char residue as high as 43.71% was found for boards incorporated
with 30% sodium aluminate (Table 2). This represented
the highest char residue amongst the treated samples. The amount of char
residue formed during pyrolysis increased from 41.46 to 43.71% for board
treated with sodium aluminate for 10 and 30% concentration levels, respectively.
It is known that wood consist of about 50% carbon and this means any char
residue values approaching 50% would give out minimum production of flammable
volatiles or smoke (Abdul Rashid and Murphy, 1993). Boards treated with
zinc borate and aluminum trihydrate also showed a good percentage of char
residue showing minimal production of flammable volatiles gas or smoke.
Flame retardant chemical could minimize production of flammable volatile
gases. Char residues create a screen on the surface of MDF against conduction
of radiated heat that eventually retard diffusion of inflammable gases
from within (Wang et al., 2004)
The TGA graphs (Fig. 2) were extrapolated from 0-550Â°C
in order to get the behavioral pattern of weight loss. The TGA curves
showed an initial small decrease in the weight of samples between 100
to 150Â°C due to the release of moisture remaining in the samples.
At 180Â°C pyrolysis began to start and becomes exothermic at 240-260Â°C
whereby condensable vapors are produced such as acetic acid, furfural
and methanol (Gao et al., 2006).
||Heat of Absorption of Flame retardant MDF compared with
the controls obtained from DSC (SA: Sodium aluminate; ATH: Aluminum
trihydrate; ZB: Zinc borate)
According to Fig. 2, thermal degradation of a board
probably started at about 180-200Â°C. Troitzsch (1998) suggested that
thermal degradation started at about the same temperature range. During
this process, aluminum oxide and water vapor are released in an endothermic
reaction. This aluminum oxide will form an insulating protective layer
on the substrate and water vapor will act as a diluting agent in the gas
phase and forms an O2 displacing protective layer over the
Thermal degradation for boards incorporated with zinc borate was closely
similar to that of the control boards. A maximum amount of char residue
of 35.65% was observed at 30% treatment levels. Flame retardant chemicals
create less flammable gases and produce more char residue and water. This
could lead to dehydration and charring of cellulose (Gao et al.,
2006). Boards incorporated with sodium aluminate seems to be the most
effective flame retardant chemical. This would be probably due is to high
percentage of char formed from samples and can be seen with as little
as those boards incorporated with 10% treatment of sodium aluminate.
It indicated that sodium aluminate has the highest heat of absorption
during decomposition compared to other flame retardant chemicals (Fig.
3). Boards incorporated with sodium aluminate showed to be the most
the effective flame retardant compared to that of aluminum trihydrate
and zinc borate. The heat of absorption could result in some amount of
water being liberated (Jian et al., 2001). Because of this, oxygen
content will be reduced due to the gas phase dilution by the water vapor
produced. In short, the mechanism of flame retardant chemical caused dehydration
of a board. This dehydration reaction will lead to the higher levels of
char and limited the amount of volatiles gas. Strong endothermic decomposition
in the DSC will reduce combustible gases and prevent access of the surface
to oxygen that might suppress ignition.
Flame retardant chemicals reduced the mechanical properties of MDF
except for modulus of rupture. The mechanical properties reduced considerably
after a cyclic test. The boards incorporated with aluminum trihydrate
gave an overall best performance in both physical and mechanical properties
followed by boards treated with sodium aluminate and zinc borate. Char
Index decreased as the proportion of flame retardant chemical increased
while weight loss was reduced as the proportion of flame retardant chemical
increased. The study indicated that boards incorporated with sodium aluminate
as the most effective flame retardant followed by boards incorporated
with aluminum trihydrate and zinc borate.
We would like to thank MOSTI, Malaysia through the IRPA program
for financial support, Hexion Specialty Chemicals for providing adhesive
samples, Borax Inc., for the zinc borate samples, Merbok MDF Sdn. Bhd.
for supplying the rubberwood fibers.