A distributed activation energy model for the pyrolysis of lignocellulosic biomass

Abstract

The pyrolysis kinetics of xylan, cellulose, and eight lignocellulosic biomass samples (including corn stover, cotton stalk, palm oil husk, pine wood, red oak, sugarcane bagasse, switchgrass, and wheat straw) were measured in a TA SDT Q600 thermogravimetric analyzer. The kinetic data of xylan and cellulose were fitted to a distributed activation energy model (DAEM) where the activation energies for the pyrolysis of each reactant followed a Gaussian distribution. The activation energy distribution peaks were centered at 178.3 and 210.0 kJ mol⁻¹ for xylan and cellulose respectively. The standard deviations of the activation energy distributions for xylan and cellulose were 5.85 and 0.94 kJ mol⁻¹. A three-parallel-DAEM-reaction model was used to fit the pyrolysis kinetics of the lignocellulosic biomass feedstocks. The model assumes that biomass contains three independent reacting pseudo-components all of which have a Gaussian distribution for their activation energies. The first, second and third pseudo-components represent the fractions of hemicellulose, cellulose and lignin, respectively. The activation energy distribution peaks for the three pseudo-components were centered at 169.7–186.8, 204.2–212.5, and 237.1–266.6 kJ mol⁻¹ for the different biomass feedstocks. The standard deviations of the activation energies were 4.6–8.8, 0.7–1.8, and 26.5–41.8 kJ mol⁻¹ for the hemicellulose, cellulose and lignin fractions respectively. This indicates that the activation energy distribution for lignin has the widest distribution, and cellulose has the narrowest distribution. The biomass compositions obtained by the fitting of the pyrolysis kinetics agreed well with published values obtained from wet chemical analysis techniques.