Zakir Khan1, Suzana Yusup1, Murni M. Ahmad1, Yoshimitsu Uemura1,

Vuoi S. Chok2, Umer Rashid1 and Abrar Inayat1

1Universiti Teknologi PETRONAS, Perak 2Platinum Energy Sdn. Bhd., Kuala Lumpur

Malaysia

1. Introduction

Malaysia is the largest producer of palm oil and contributes 43% of worldwide production (Shuit et al., 2009). Beside palm oil, palm oil industry generated 169.72 million metric tons solid wastes which contribute 85.5% of total biomass waste produced in the country (Khan et al., 2010). This huge amount of wastes can be converted into valuable chemical feed stocks and fuels due to environmental problems associated with conventional fossil fuels.

It is well known that lignocellulosic biomass mainly consists of hemicellulose, cellulose and lignin. The usual proportions (wt%) vary as 40-50% cellulose, 20-60% hemicellulose and 10­25% lignin (Yang et al., 2007). The thermal decomposition of these individuals is important since they influence the basics of thermochemical conversion processes such as pyrolysis, combustion and gasification. Decomposition of these components is intensively studied in the literature. Demirbas et al. (2001) observed the ease of lignocellulosic biomass components decomposition as hemicellulose > cellulose >>> lignin. Based on different reasoning, Yang et al. (2007) proposed different decomposition regions of 220-300 °C, 300­340 °C and >340 °C for hemicellulose, cellulose and lignin, respectively. Lignin is the last to decompose due to its heavy cross linked structure (Guo & Lua, 2001).

Several techniques are available to study the kinetics of biomass decomposition. Among these, thermogravimetric analysis (TGA) is the most popular and simplest technique (Luangkiattikhun et al., 2008), based on the observation of sample mass loss against time or temperature at a specific heating rate. TGA provides high precision (Varhegyi et al., 2009), fast rate data collection and high repeatability (Yang et al., 2004) under well defined kinetic control region.

Very few attempts have been carried out to study the kinetics of empty fruit bunch (EFB) and palm shell (PS) using TGA. Guo & Lua (2001) presented the effect of sample particle size and heating rate on pyrolysis process and kinetic parameters for PS. They concluded a first order reaction mechanism for the decomposition of PS at different heating rates. They also suggested higher heating rates for faster and easy thermal decomposition of PS. Yang et al. (2004) studied activation energy for decompositions of hemicellulose and cellulose in EFB and PS by considering different temperature region for first order kinetic reaction. They evaluated average activation energy and pre-exponential factor from single-step decompositions of hemicellulose and cellulose. Luangkiattikhun et al. (2008) considered the

effect of heating rate and sample particle size on the thermogram behaviour and kinetic parameters for palm oil shell, fibre and kernel. They observed that there is no significant effect of particle size on the thermogram behaviour at lower temperature i. e. <320 °C for palm oil shell. They further proposed nth order reaction mechanism to evaluate the kinetic parameters based on different models.

Previous works reported on EFB and PS kinetics were based on single heating rate in which activation energy is only a function of temperature. The present work evaluate the kinetic parameters based on a method, which requires at least three sets of experimental data generated at different heating rates. This method allows the dependence of activation energy on temperature and conversion at a desired heating rate (Vyazovkin & Wight, 1999) Secondly, lignin decomposition in EFB and PS is not intensively studied at relatively high heating rates. Present work considers lignin decomposition in EFB and PS to understand the effect of lignin content on kinetic parameters and decomposition rate. Furthermore, pure lignin decomposition is studied based on its thermogram analysis and kinetic parameters.

In this work, the kinetics of biomass decomposition which includes EFB, PS, pure cellulose and lignin were investigated using TGA under non-isothermal conditions. The detail thermogram analysis was presented to understand the decomposition of cellulose, hemicellulose and lignin as major components in lignocellulosic biomass. The decomposition kinetics of cellulose and lignin were studied under single-step first order kinetic model. Meanwhile, the decomposition of EFB and PS were reported based on single­step nth order kinetic model. Activation energy, pre-exponential factor and order of reaction were determined and discussed in comparison to the values reported in the literature.