The first step consisted in a field investigation carried out in Waste Electric and Electronic Equipment (WEEE) treatment plants coupled with a bibliographic product analysis and FT-IR spectroscopy polymers identification. Three main polymers of the thermoplastic fraction of small appliances were identified: in the external cases it was possible to find acrylonitrile- butadiene-styrene (ABS) and high impact polystyrene (HIPS), while polybutadiene terephthalate (PBT) was contained in the printed circuit boards (PCBs). Taking into account this identification, a ternary polymer mixture of ABS-HIPS-PBT was prepared as a representative sample of the thermoplastic fraction contained in WEEE (real WEEE sample). From the thermal characterization (proximate and ultimate analysis, high heating value (HHV) direct measurement and Energy-Dispersive-X-Ray-Fluorescence analysis (ED-XRF)) the only polymer whose properties sensibly differ from the analogous virgin polymer is the one contained in PCBs. A kinetic analysis of pyrolysis occurring on the three components and on their ternary mixture was performed using a thermogravimetry (TG) apparatus in argon atmosphere under non isothermal conditions. Triplicates of TG experiments at four heating rates of 2, 5, 10 and 15 K min-1 were carried out and two different model-free approaches were adopted, namely the Kissinger and Ozawa-Flynn-Wall methods in order to determine the activation energy E (as a single mean value or as a function of the degree of conversion α). The conversion dependencies of both activation energy and pre-exponential factors were determined as well as the reaction model, representing the reaction mechanism. The suitability of the models selected was tested using the Akaike's Information Criteria (AIC) score, being the geometric model R3 the best found for pyrolysis of ABS, HIPS and real WEEE samples, while PBT sample showed an uncertainty between the R3 and the diffusion D2 model. The reaction time values to achieve the maximum pyrolysis rate in the three main components and in the real WEEE sample were also calculated. © 2014 Elsevier Ltd. All rights reserved.

cited By 6