Co-pyrolysis behaviour of de-oiled mahua cake and waste LDPE using thermogravimetric analysis and artificial neural network

Pyrolysis of biomass-plastic waste mixtures offers a promising pathway for sustainable energy recovery, yet the underlying kinetic interactions remain largely unexplored. This study investigates the co-pyrolysis behavior of de-oiled mahua cake (DOMC) and waste low-density polyethylene (LDPE) using thermogravimetric analysis (TGA), model-free kinetic modeling, and artificial neural network (ANN) prediction. Kinetic analysis using Flynn-Wall-Ozawa (FWO), Kissinger-Akahira-Sunose (KAS), Starink, Tang, and Boswell methods revealed a reduction in activation energy (E_a) for LDPE when mixed with DOMC, suggesting a synergistic effect that enhances decomposition efficiency. The ANN model demonstrated high predictive accuracy (R² ∼1), effectively capturing pyrolysis behavior across different heating rates (5, 10, and 20 °C/min). The findings highlight the potential of co-pyrolysis for reducing energy barriers in plastic waste degradation and underscore the applicability of AI-based predictive modeling for pyrolysis optimization.