The oil palm fibre used in the fabrication of the composite fibreboard was sourced within the University of Nigeria, Nsukka community. Flame retardants fillers consisting of aluminium tri-hydroxide (Al₂(OH)₃) of particle size 10µm and ammonium polyphosphate ((NH4PO3)n(OH)₂) a white-free flowing powder soluble in g/100ml of H₂O with average particle size of 15µm, n-Hexane, litmus paper, distilled water and the polyester resin used were supplied by Joe Chem ventures. Gum Arabic purchased in the Northern part of Nigeria was further processed into a fine powder of particle size 300 µm used as a binder and source of carbohydrate to formulate new flame retardant specie. Unsaturated polyester resin (UPR) was cured with 2% methyl ethyl ketone peroxide (MEKP) as catalyst and 1% cobalt (Co) as accelerator. All chemicals were used without further purification. Note that the exact chemical structure of UPR was note supplied by the manufacturer.

Oil palm fibre (OPF) as received (Figure 1a) were first extracted by washing with hot water to remove the remaining residual oil retained during oil extraction, a method proposed by Vijaya et al (25) and then soaked in n-Hexane overnight to complete leaching and further remove impurities. The fibres were then treated with 5% (NaOH) solution for about 2hrs to avoid fibre damage as displayed in (Figure 1b) to improve the compatibility with the polyester resin. Afterwards, the fibres were washed with distilled water until blue litmus paper turned red which indicates that excess concentration of NaOH have been neutralized and then sun dried for 3 days to remove moisture content as shown in (Figure 1c). The ready to use fibres in (Figure 1d) were used to fabricate the fibreboard using hand lay-up compression moulding technique as shown in (Figure 1e). The required quantities of the reinforcement and polyester resin used to produce the fibreboards was obtained using mass fraction model (26) through equations (Eq. [1] to [3]) as shown below. Four (4) fibreboards were produced and investigated (Figure 1f). Table 1 illustrates the composition of the various oil palm fibre composite (OPFC) fibreboard produced including the percentages of the added flame retardants.

Where, X is the Mass of the various constituents x, y, and z and the mass fraction can be obtained through the relationship denoted in Eq.[2] and Eq. [3].

or

V_f = volume fraction of fibre, M_f = mass fraction of fibre, r_c = density of the various constituents x, y and z (g/cm³), r_f = density of the fiber and m_f, _f2, m = measured mass of fibre (s) and matrix respectively.

Izod impact test was performed on the fabricated composite fibreboard specimens using an impact tester in accordance with ASTM D 256 standards. From the fibreboard the IS dimension (100x10x10) mm was obtained. Prior to mounting on the test machine, the test specimen was notched to a depth of 2mm with a v-shaped hand file. The notched test specimen was then mounted on the impact-testing machine, which is operated to impact a blow to fracture the specimen at the opposite end of the notch by releasing the suspended handle of the pendulum swing. The impact strength measured as the absorbed before fracture was then read off on the calibrated scale. The specimens were repeated three (3) times and an average absorbed energy value recorded.

TGA analysis of fibreboard was conducted using thermos-gravimetric analysis (TGA/DSC 1; Mettler Toledo, UKBRC, Edinburgh UK). A 5mg samples were first heated for 10min at 105⁰C under nitrogen gas (N₂) to determine moisture content; the temperature was then raised at 25⁰C min^-1 to 900⁰C where it remained for a further 10min to determine volatile matter content. Finally, air was introduced to the system combusting the sample (also at 900⁰C) for 20 minutes in order to determine the ash content.

The cone calorimeter apparatus was used to study the flammability (fire reaction properties) of the fabricated OPFC fibreboard according to ASTM E 1354. The specimens (100mm x 100mm x10mm) cut from the fibreboard were wrapped in aluminium foil; along the side and bottom to reduce heat losses as specified in the standard. The specimens were exposed to 50kwm^-2 heat flux horizontal orientation. During testing, the time to ignition, heat release rate, mass loss rate, residual mass, smoke and toxic emissions were obtained. The test specimens were repeated severally and a sensible result was adopted for this study.

The results obtained for the impact strength as shown in (Figure 2) reveals that the absorbed energy of the OPFC fibreboard with FR did not show any improvement. In fact, the addition of OPFC_12%ATH and OPFC_{18%ATH/APP-GAP} in the fibreboard slightly deteriorated the amount of absorbed energy from 71.9kJ/m² to 65.4kJ/m² (9% decrease) compared to the OPFC_0%. A similar trend in decrease has also been reported by Redwan et al (13) using different ATH content empty fruit reinforced/middle density board. It was however observed that the addition of OPFC_12%APP-GAP in the fibreboard maintained the same absorbed energy with the OPFC_0%. This results agrees with the IS report by Nikmatin et al (27). In this study, the presence of GAP probably played a significant role in enhancing the fibreboard compactness and thus facilitated the transfer of stresses between the fibre, PR and the FRs during blow. The reason for the decrease in impact could be a non-uniform dispersion coupled with agglomerated FR particles which provides locations of stress concentrations, thus provides sites for crack initiations as reported by Subasinghe et al (28).

In Table 2, the parameters describing the thermal response of the OPFC are presented which was derived from the thermogravimetric (TGA) and derivative of TG (DTG) curves as shown in Figure 3. From the sigmoidal shaped curves, it reveals that the OPFC is typical single weight loss step degradation. The initial degradation process of the fibreboards was noted at 200⁰C, indicative of the loss of water vapour from the fibres as well as confirms the hydrophilic nature of the fibres (29, 30). As the heating continues the fibreboard began to lose weight until it reaches the actual degradation of the fibreboards around the onset decomposition temperature (T₀) before degrading sharply noted at the peak DTG. The overall degradation process reveals that the OPFC_12%ATH addition in the fibreboard exhibited a better thermal stability as the T₀ reached 376⁰C from 371⁰C and the peak DTG reached 421.93⁰C from 409.82⁰C compared to the OPFC_0%. Similar reports for the increase in thermal stability with ATH in oil palm empty fruit fibre reinforced epoxy was found in the work of Khalili et al (7). The increase in thermal stability agrees with the findings of Happarachchi (31), attributed to the endothermic dehydration and the subsequent release of 35% of H₂O crystallization into the gas phase which led to the formation of a ceramic layer of γ-Al₂O₃. It was observed that the OPFC_{18%ATH/AP-GAP} hybrid formulations gained in weight loss and char residue which stood at 83.3% and 15.46% compared to the OPFC_0% at 92.7% and 6.15% respectively. This suggests a slower decomposition and significantly enhanced thermal stability as well as a superior advantage in terms of their flame retardancy. ATH and APP upon reaction could be responsible for the high char formation as reported in the work of (32).

The fire risk of the fibreboard was assessed through its time to ignition (Tig) response which is responsible for the fibreboard flame spread. Table 3, shows the various ignition times obtained in the cone calorimeter at 50kWm^-2. It reveals that the OPFC_12%ATH and OPFC_{18%ATH/APP-GAP} fibreboards delayed longer the release of combustible volatiles. The combined effect of tri-hydroxide (ATH) and ammonium polyphosphate and Gum Arabic powder (APP-GAP) could be responsible for the delay in Tig which is attributed to strictly the thermal decomposition of ATH into (Al₂O₃ and H₂O) since water vapour dilutes the volatile gases being formed. This may have cooled the gases and delay Tig. Another reason could be that APP-GAP dehydrates the fibreboard upon heating to form char that prevents the release of combustible volatiles.

The HRR profile of the OPFC obtained from cone calorimeter are depicted in (Figure 4). The heat release rate (HRR) is a key property in evaluating the fibreboards limits to fire safety. From the heat release rate curves it is clearly seen that shortly after ignition, a sharp rise and then a sudden decline was observed for all the fibreboard types which indicate the activities of the combustible products during combustion. It was observed that the OPFC_0% exhibited a double peak HRR, indicative of a high release of combustible rate probably caused by cracks in the char structure of the underlying composite fibreboard substrate as reported by Schartel and Braun (33). When flame retardant was added to the fibreboard it reveals a broader appearance which stayed at a lower profile throughout the burning process, indicative of the flame retardant interaction with the combustible products. In (Table 3) it was found that the addition of OPFC_{18%ATH/AP-GAP} followed by OPFC_12%ATH flame retardants in the fibreboard exhibited an outstanding performance as the peak HRR and average HRR indicating the intensity of fire and the contribution to sustained fire respectively were remarkably reduced to 86.6kWm^-2 and 136.5kWm^-2 from 265.5kWm^-2 and 55.8kWm^-2 and 67.6kWm^-2 from 150.2kWm^-2 respectively compared to OPFC_0%. This could be elucidated by the effect of ATH and APP-GAP characteristic mechanism. Aluminium tri-hydroxide acts to trap the formed combustible products during the burning process, releasing water during dehydration and restricting access of oxygen to the fibreboard substrate. This agrees with the HRR profile found in the report of Nikolaeva and Karki et al (34). On the other hand, APP-GAP inhibits the mass and volatile transfer between the condense phase and gas phase once a char layer is formed, thus it contributes to the reduction in heat release rate.

The MLR of the fibreboard which is a measure of the dehydration reactions and pyrolysis was examined and presented as shown in (Table 3). It can be deduced from the specific MLR on average (SMLRavg) that OPFC_12%ATH followed by OPFC_{18%ATH/APP-GAP} were the lowest at 6.9 and 7.1 gs^-1s^-2 respectively,indicative of slow decomposing fibreboard compared to OPFC_0% which led to an enhanced residual mass (Rm) and suggest a good flame retardancy. In fact, the greater the decrease in SMLRavg, the better the reduction in the peak HRR was observed, this has also been confirmed by other authors (34-36). The combustion residue displayed for the fibreboards on the right hand side of (Figure 4) after combustion further reveals that OPFC_12%ATH left a mass of smoother black char and thin cracks indicating a better restrain of the combustion volatiles while the OPFC_{18%ATH/APP-GAP} with a fully covered, uniformly coherent and compact char residue left a mixture of gray and black mass of char indicating a rich carbonaceous char formation and the reason it attained the lowest peak HRR (86.6kWm^-2) and higher amount of Rm (54.6%). This implies that both mass and heat transfer between condense phase and gas phase was restricted and consequently the underlying material protected from further combustion of the polymer pyrolysis.

The importance of the smoke release accompanied by toxic gases is a critical aspect of human survival during a fire as inhalation is one of the greatest hazards to life; hence the knowledge of the smoke and gas release of the fibreboards is imperative. In (Figure 5), the smoke production rate (SPR) of the fibreboards as a function of time exhibits similar characteristics trend with those of the HRR curves and suggests that the release of smoke accompanied with toxic gases (Carbon-monoxide production, COP) as heat is evolved into the atmosphere. In (Table 4), it was observed that among the studied flame retardants in the fibreboard, the addition of OPFC_{18%ATH/APP-GAP} in the fibreboard caused a positive impact on the smoke and gas properties. OPFC_{18%ATH/APP-GAP} maintained the same SPRavg with the OPFC_0% at 0.05m²s^-1 while the total smoke released (TSR) was suppressed to 2447.5m²m^-2 from 3604.8m²m^-2 (32% decreases). From the carbon-monoxide production data, it shows that the entire flame retardants presence in the fibreboard suppressed the COP with OPFC_12%APP-GAP the least at 0.029Kg/Kg followed by OPFC_{18%ATH/APP-GAP} at 0.35Kg/Kg from 0.069Kg/Kg representing 58% and 49.2% respectively. The fibreboards however show that the av-SEA did not improve the entire FR studied. This could be attributed to the longer burning time of the FR fibreboard compared to the OPFC_0%.