Pyrolysis behaviour of ultrafiltration polymer composite membranes (PSF/ PET): Kinetic, thermodynamic, prediction modelling using artificial neural network and volatile product analysis

This study aims to explore the feasibility of managing ultrafiltration polymer composite membranes (UPCM) waste and converting it into valuable chemicals and energy products using a pyrolysis process. The thermal decomposition experiments were performed on polysulfone (PSF)/polyethylene terephthalate (PET) membranes using thermogravimetric analysis (TG). The vapors generated during the thermochemical process were analyzed under different heating rate conditions using TG-FTIR and GC/MS. In addition, the kinetic and thermodynamic parameters of the pyrolysis process were determined using conventional modeling methods and artificial neural network (ANN) method. The results demonstrated that the PSF/PET feedstock exhibits ahigh volatile matter content (77 % wt.%), which can be completely decomposed up to 600 degrees C by 79 wt%. While TG-FTIR analysis showed that the released vapors contained aromatic groups and benzoic acid (89.21 wt% at 15 & ring;C/min) as the main GC/MS compound. Moreover, the kinetic analysis demonstrated complete decomposition of the membranes at a lower activation energy (151 kJ/mol). Meanwhile, the ANN model exhibited high performance in predicting the degradation stages of PSF/PET membranes under unknown heating conditions. This approach shows potential for modeling the thermal decomposition of ultrafiltration composite membranes more broadly.