Pyrolysis behavior of non-textile components (buttons) and their kinetic analysis using artificial neural network

This research aims to study the pyrolysis behavior of old buttons (main non-textile components) and their kinetic behavior to convert them into energy and their original chemical compounds. The pyrolysis experiments were performed using a thermogravimetric analyzer (TG) on buttons have different composition that were defined using FTIR, elemental and proximate analysis. The composition of the valuable chemicals generated from the pyrolysis process were observed TG-FTIR and GC/MS. The kinetic parameters of the decomposition process were also studied using conventional modeling methods and artificial neural network (ANN) as an advanced machine learning tool. The results showed that polyester, nylon and their blends are the most commonly used materials in button manufacturing. The physical analysis showed that the buttons are very rich in volatile matter content (92.08–99.67 wt%) and completely decompose up to 490 °C at 92–100 wt%. Meanwhile, GC/MS showed that the pyrolysis vapors released from polyester buttons were rich in styrene (84.54 %), while caprolactam (40.30 %) was the dominant compound in nylon buttons versus naphthalene, 1,2,3,4-tetrahydro-2-phenyl- (67.71 %) was the major compound in the mixture sample. The kinetic analysis showed that the activation energy of the degradation process was in the ranges of 152–202 kJ/mol (polyester), 156–201 kJ/mol (nylon), 402–449 kJ/mol (mixed) and the ANN model was successfully trained and predicted the degradation regions of the buttons. Accordingly, pyrolysis of buttons is highly recommended to valorize buttons and convert them into parent chemical compounds.