Key points are not available for this paper at this time.
This study investigates the pyrolytic behaviour of lignin using an integrated kinetic and thermodynamic framework based on thermogravimetric data. Four iso -conversional models-Friedman, Ozawa-Flynn-Wall (OFW), Kissinger-Akahira-Sunose (KAS), and Starink-were applied to monitor how the activation energy ( Ea α ) evolves throughout the conversion ( α = 0.1–0.9) range. The non-uniform activation energy ( Ea ) pattern confirmed that lignin does not decompose through a single mechanism but undergoes a multi-stage transformation. Among the evaluated methods, KAS and Starink generated the most consistent Ea profiles. The associated pre-exponential factor ( A ) and thermodynamic parameters- enthalpy ( Δ H ), Gibbs free energy ( Δ G ), and entropy ( Δ S ) further indicated that pyrolysis of lignin requires substantial energy input and proceeds in a non-spontaneous manner. A generalized masterplot interpretation revealed a clear shift in the dominant reaction mechanisms with conversion, transitioning sequentially through diffusion-controlled (D1), first-order (F1), and phase-boundary-controlled (R2) regimes, highlighting the coexistence of multiple kinetic pathways and the progressively evolving nature of lignin degradation. To complement the kinetic evaluation, an artificial neural network (ANN) was developed using temperature and heating rate as inputs and conversion as output to predict conversion behaviour. The ANN closely replicated the experimental conversion curves, demonstrating strong capability in capturing the nonlinear decomposition pattern with high predictive accuracy ( R 2 = 0.9998, mean square error = 2.2364 × 10 −6 ) and reducing reliance on extensive experimental iterations. By coupling iso -conversional kinetic modelling with ANN-based prediction, this work delivers the first hybrid modelling framework specifically tailored to lignin derived from mixed-source industrial black liquor.
Meena et al. (Sun,) studied this question.