A comprehensive study of facemasks pyrolysis using Py-GC/MS, kinetic analysis and ANN modeling

[Display omitted] The thermo-kinetics of pyrolysis and product distribution of facemasks, as a blend of filter layers were explored in this study. Pyrolysis products were studied using Py-GC/MS at 550 °C for 30 s, resulting in predominantly aliphatic hydrocarbons (82.6%): alkanes (34.5%) and alkenes (48.1%). Notable gaseous products identified include propene, 2-methyl pentane, and 2,3-dimethyl-1-pentene, while the dominant species among the cycloalkanes were 1,2,3,5-tetraisopropylcyclohexane and 1,3,5-trimethylcyclohexane. Furthermore, we developed a chemical reaction mechanism to describe the main products formed during pyrolysis. Besides, the activation energy was predicted using model-free methods namely FR (214.2 kJ/mol), KAS (200.5 kJ/mol), and FWO (200.6 kJ/mol), as a function of conversion. The Coats – Redfern (CR) model-fitting method revealed that the pyrolysis reaction mechanism within the temperature range of 400 – 550 °C (pyrolysis active zone) belonged to one-dimensional diffusion and contracting cylinder model. The reliability of these results was further affirmed using the Criado method, showing agreement between the experimental and theoretical master plots. The thermodynamic parameters for facemask degradation indicated an endothermic process (ΔH = 205.5 kJ/mol, ΔG = 182.4 kJ/mol, and ΔS = 0.03 kJ/mol·K). To predict weight loss during facemask pyrolysis, we developed an artificial neural network (ANN) model that considered heating rate and temperature as inputs. The most efficient model structure involved an ANN with 2 input layers (2 neurons each), 2 hidden layers (each with 10 neurons), and an output layer (1 neuron). This study is crucial for advancing our understanding of the theoretical aspects of polymeric waste pyrolysis.