Loading...

Kinetic Investigation of the Thermal Treatment of E-Waste with Particular Emphasis on Printed Circuit Boards (PCBs)

Shokri, Ali | 2022

175 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 55523 (06)
  4. University: Sharif University of Technology
  5. Department: Chemical and Petroleum Engineering
  6. Advisor(s): Fotovat, Farzam
  7. Abstract:
  8. In this study, pyrolysis features and corresponding kinetic behavior of non-metallic fractions of the paper-laminated phenol resin printed circuit boards (P-PCB) and fiberglass-reinforced epoxy resin printed circuit boards (E-PCB) were investigated by the use of simultaneous thermogravimetric analysis (TGA/DTG). The kinetic parameters of the P-PCB abs E-PCB during pyrolysis process were estimated from the experimental data by means of three iso-conversional kinetic models, including Friedman, Flynn-Wall-Ozawa (FWO), and Kissinger-Akahira-Sunose (KAS) methods. The activation energy range obtained for P-PCB and E-PCB utilizing the Friedman method was within 117–594.1 and 88.42-295.3 kJ.mol-1, from FWO method was within 138.2-529 and 108.7-234.7 kJ.mol-1, whereas KAS method demonstrated range of 137.1–543.1 kJ.mol-1 and 107.6–238.6 kJ.mol-1, respectively. Furthermore, discrete distributed activation energy models (DAEMs) with first-order reaction mechanism has been developed to describe the pyrolysis kinetics of P-PCB and E-PCB during the pyrolysis process. According to the kinetic parameters obtained by applying the discrete DAEM algorithm, the thermal decomposition process of P-PCB and E-PCB can be accurately modelled with 16 and 6 dominating parallel first-order reactions, respectively. In addition, highly efficient artificial neural network (ANN) models were developed and trained to predict the thermal behavior of P-PCB and E-PCB as a function of heating rate and temperature. The effect of important parameters such as the number of layers, the number of neurons, the type of activation functions, and the training algorithm on the neural networks efficiency was investigated to optimize the network topology. The optimization studies demonstrated that for thermal degradation of P-PCB and E-PCB the best performance was achieved with ANN that had 5-5 and 6-7 neurons with tansig-logsig non-linear activation function combination.
  9. Keywords:
  10. Pyrolysis ; Artificial Neural Network ; Kinetics Modeling ; Thermal Gravimetric Analysis ; Fiberglass-Reinforced Epoxy Resin Printed Circuit Board ; Paper-Laminated Phenolic Printed Circuit Board ; Distributed Activation Energy Model (DAEM)

 Digital Object List

 Bookmark

No TOC