Pyrolysis behavior of Sabina chinensis leaves: kinetic and thermodynamic analysis based on TG-FTIR and BP neural network

Abstract

Biomass resources play a crucial role in meeting our energy, chemical and material demands. Sabina chinensis, a biomass material with diverse constituents, can be effectively treated using pyrolysis, enhancing its resource utilization. This study investigated the pyrolysis behavior of Sabina chinensis leaves, whose pyrolysis process exhibited three distinct stages: (1) stage Ⅰ (moisture evolution), (2) stage Ⅱ (decomposition of hemicellulose, cellulose, and lignin, yielding CO₂, methane, methanol, formaldehyde, and other products), and (3) stage Ⅲ (gradual decomposition of metal carbonates in residues). Using deconvolution and model-free methods, fundamental kinetic parameters for the pyrolysis of hemicellulose, cellulose, and lignin were estimated. Activation energies of 160.70, 180.49, and 159.73 kJ/mol were obtained for hemicellulose, cellulose and lignin, respectively. Thermodynamic analysis revealed enthalpy change values of 153.74–181.29 kJ/mol, Gibbs free energy change values of 277.94–338.10 kJ/mol, and negative entropy change values, demonstrating the energy potential of Sabina chinensis leaves. Furthermore, a backpropagation neural network, trained using mass and gas product data at multiple heating rates, successfully predicted the temperature-dependent mass loss and gas yield evolution throughout the pyrolysis process.