Pre-treatment of oil palm fronds by ozonolysis for levulinic acid production

Abstract

Oil palm fronds (OPF) is an attractive feedstock for levulinic acid (LA) production due to its availability and high content of cellulose, but requires a pre-treatment because of its heterogeneity structure. This study explored the potential of ozonolysis pre-treatment of OPF for LA production. The suitable design of experiment and response surface methodology (RSM) by Statistica software ver. 8.0 was employed to study the effect of process parameter and to determine the optimum condition. The lignin degradation and total reducing sugar (TRS) recovery were set as responses while ozonolysis pre-treatment condition i.e. particle size, moisture content, ozone flowrate, reaction time, ozone concentration and part of OPF were set as independent parameters. The multi-response optimization of lignin degradation and TRS recovery for ozonolysis pre-treatment were verified by Box-Behnken design with four selected independent process parameters. The physico-chemical properties of OPF and treated OPF (ODT) were analysed by thermal gravimetric analysis (TGA), Fourier transform infra-red (FTIR), X-ray diffractogram (XRD), N2-adsorption, scanning electron microscopy (SEM) and field emission scanning electron microscopy with energy dispersive X-ray spectroscopy (FESEM-EDX). LA was produced by conventional acid hydrolysis. An optimal region of study for lignin degradation was recommended at 25–40 wt.% moisture content, particle size bigger than 0.6 mm and ozone flow rate faster than 70 mL/min within 60 min. The TRS recovery is independent of lignin degradation. 75.8 % of TRS recovery of ODT was attained at 0.63 mm, 30 wt.%, and 60 mL/min compared to 46.7 % of OPF. FESEM and SEM depicted that the cell wall of OPF was broken, exposing the microfibril and cellulose rosette structure during the pre-treatment. Rising crystallinity index from 36.1 % (OPF) to 44.7% (ODT) from XRD confirmed the removal of amorphous lignin and hemicellulose component as shown by TGA and FTIR analyses. The decreasing surface area does not hinder the subsequence hydrolysis reaction; reducing crystal size up to 60.5%, and increasing pore diameter and volume gave advantage for the reaction. The particle size-moisture content interaction is important for lignin degradation while the moisture content-reaction time interaction is crucial for the TRS recovery. Larger OPF particle size increases lignin degradation and TRS recovery due to interfacial surface tension. The reaction and mass transfer in water film was controlled by moisture content and reaction time. The optimum lignin degradation (84.7 wt.%) and TRS recovery (99.9 %) were reached at 0.8 mm particle size, 40 wt.% moisture content, 75 min reaction time and 105 mL/min ozone flow rate with 19.5 % ozone consumption. The LA yield of ODT at 180 ºC for 1 h and 4 wt. % H2SO4 increased up to 4.72 times than OPF and comparable to commercial microcrystalline cellulose. 8.7 wt.% of LA recovery was attained by ozonolysis pre-treatment. The findings from this study provide the insight background of the ozonolysis pre-treatment of OPF for the further stage of commercialization that could contribute to Malaysia’s economy.