Wang, Ziwei2022-06-082022-06-082022-01https://hdl.handle.net/11299/227927University of Minnesota Ph.D. dissertation. January 2022. Major: Chemical Engineering. Advisor: Matthew Neurock. 1 computer file (PDF); xix, 171 pages.Fast pyrolysis offers one of the most accessible ways to convert macromolecular feedstocks, such as biomass or plastic waste, into small molecules that can subsequently be used to produce liquid fuels and chemicals. The complexity of the feedstock and the dynamic changes in condensed phase chemical environments make it challenging to elucidate elementary kinetics and reaction mechanisms and control the selectivity to products. These effects also manifest in macroscopic phenomena such as measurable differences in kinetics and product distribution with changes in reaction temperature, feedstock composition, and even pellet size of the feedstock. These complexities are ultimately dictated by molecular transformations and controlled by the molecular structure and local chemical environments. Understanding the molecular processes in pyrolysis is crucial for the development of large-scale and economical biomass conversion or plastic upcycling facilities. This dissertation presents the development and applications of an ab initio-based kinetic Monte Carlo and Molecular Dynamics (KMC+MD) simulation approach that can model the kinetics of the fast pyrolysis and thermal reaction networks for biomass and polyolefin plastic feedstocks. This approach tracks detailed atomic structural information, including 3D coordinates of atoms and connectivity of chemical bonds in the feedstock molecules. It uses a stochastic simulation algorithm (SSA) to track and carry out the elementary reaction steps and uses classical MD simulations to follow the dynamics of the feedstock and the reaction environment as reactions proceed. The elementary step kinetics for the KMC simulations were established from detailed first-principle density functional theory (DFT) calculations. The detailed atomic-structural information is retained throughout the simulation, thus allowing the simulations to follow molecular transformations and the local environment during the reaction. As such, the simulations capture the unique kinetic manifestations that would otherwise be lost in composition-based deterministic models. This KMC+MD simulation approach is first used to model the pyrolysis reaction pathways of cellulose, including the paths to form levoglucosan as well as light oxygenates. The simulation results are able to reproduce temporal experimental product distributions and, in addition, gain molecular-level insights into the unique catalytic features that control the kinetics. The generality of the KMC simulation framework allows it to readily be adapted and used to simulate the kinetics and product distributions of polyolefin pyrolysis, including polyethylene and polypropylene feedstocks. The simulated polyolefin pyrolysis results are extensively compared with experimental data reported in the literature and those obtained by experimental collaborators in Prof. Paul Dauenhauer’s group. More generally, the simulation framework presented in this dissertation provides a powerful tool to study the thermal degradation of different polymeric feedstocks in pyrolysis systems that challenges the capability of deterministic models. The simulation approach offers molecular-level mechanistic insights and has direct applications in reactor modeling and techno-economic analyses of large-scale pyrolysis facilities.enbiomasskinetic modelingkinetic Monte CarlopolyolefinpyrolysisThesis or Dissertation