1.1 Materials preparation and experimental procedure

Cellulose in fibrous powder form and lignin in brown alkali powder form were purchased from Sigma Aldrich Sdn. Bhd., Malaysia. EFB and PS were collected from local palm oil industry in Perak, Malaysia. Biomass samples were dried at 105°C and the weighted was monitored at one hr interval, until the readings became constant. Samples were then grinded to particle size of 150-250pm. The method for drying, characterization and analysis were given in previous work (Abdullah et al., 2010). The biomass, pure cellulose and lignin properties are given in Tables 1 and 2.

The biomass decomposition experiments were carried out in EXSTAR TG/DTA 6300 (SII, Japan). N2 was used as inert gas with a constant flow rate of 100 ml/min for the entire range of experiments. The sample initial weight used in all experiments was within the range of 3­6 mg. TG experiments were performed at heating rate of 10, 30 and 50 °С/ min. All samples were first heated from 50 °C to 150 °C where it was kept constant for 10 min to remove moisture content, and then heated up to the final temperature of 800 °С. All experiments were carried twice for reproducibility. No significant variations were observed in the second experimental measurements.

1.2 Kinetic parameters determination

The biomass decomposition rate under non-isothermal condition is described (Cai & Bi, 2009).

A E

—-exp(—— ) f (a)

в RT
f (a) depends on the reaction mechanism as listed in Table 3 and a is the mass fraction reacted.

w — w a = —°

w0 — wf

Where, — and f shows initial and final sample weight

“ da AT E AE E

g(“) = f—— = — f exp(—- )dT = p(——- )

{ f (“) в { RT $R RT

p(E/RT) function has no exact analytical solution, and therefore different approximations are reported to evaluate the function (Budrugeac et al., 2000). The method developed by Flynn-Wall (Flynn & Wall, 1996), Ozawa (1965) using Doyle’s approximation (1961) is the most popular and commonly used by several researchers for biomass decomposition (Cai & Bi, 2009; Hu et al., 2007; Zhouling et al., 2009).

E

E -1.0516(- )

p(— ) = 0.0048e RT

RT

(3) is then rearranged for в

Іпв = ln( A“E“ ) — 5.331 -1.0516—E^ (5)

R g(“) RTV V ‘

Where; w = sample weight (mg); в = heating rate (K/ min); R = universal gas constant (8.314 x 10-3 kJ mol-1 K-1)

To determine the activation energy, lnei vs. 1/Ta, i is plotted for different a values and heating rates (i) to give a straight line and the slope of which gives the activation energy (Doyle, 1961; Ozawa, 1965; Zsako, & Zsako, 1980; Flynn & Wall, 1996).

Analysis	EFB	PS
Volatiles	84.61	81.03
Ash	5.50	4.10
Fixed Carbon (by difference)	9.89	14.87
C	40.73	49.65
H	5.75	6.13
N	1.40	0.41
S	0.22	0.48
O (by difference)	51.90	43.33
Cellulose®	38.30	20.8
Hemicellulose a	35.30	22.7
Lignina	22.10	50.7

Table 1. Biomass analysis (wt% dry basis)

a Kelly-Yong et al. (2007)

Analysis	Сєі^^є	Lignin
С	43.09	47.71
H	5.96	4.53
N	0.13	0.04
S	0.14	4.24
O (by difference)	50.67	43.40

Table 2. Pure cellulose and Lignin properties (wt % dry ash free)

Function	f(a)	g(a)
First order reaction	1-a	-ln(1-a)
Second order reaction	(1-а)2	(1-а)-1-1
Third order reaction	1/2(1-а)3	(1-а)-2-1
nth order reaction	(1-a)	1-(1-a)1-n/1-n

Table 3. Different f(a) and g(a) values based on kinetic control regime (Ahmad et al., 2009)

1.3 Model for kinetic parameter determination

The following assumptions are considered for the decomposition of EFB, PS, pure cellulose and lignin.

• Reaction is purely kinetic controlled.

• The decompositions follow single-step processes.

• First order reaction kinetics is considered for pure cellulose and lignin and PS and EFB kinetics are assumed to be nth order.

• No secondary reaction takes place among the gaseous products.