Computational Modeling of Soot Nucleation.

Abstract

Recent studies indicate that soot is the second most significant driver of climate change - behind CO2, but ahead of methane - and increased levels of soot particles in the air are linked to health hazards such as heart disease and lung cancer. Within the soot formation process, soot nucleation is the least understood step, and current experimental findings are still limited. This thesis presents computational modeling studies of the major pathways of the soot nucleation process. In this study, two regimes of soot nucleation – chemical growth and physical agglomeration – were evaluated and the results demonstrated that combustion conditions determine the relative importance of these two routes. Also, the dimerization process of polycyclic aromatic hydrocarbons, which has been regarded as one of the most important physical agglomeration processes in soot formation, was carefully examined with a new method for obtaining the nucleation rate using molecular dynamics simulation. The results indicate that the role of pyrene dimerization, which is the commonly accepted model, is expected to be highly dependent on various flame temperature conditions and may not be a key step in the soot nucleation process. An additional pathway, coronene dimerization in this case, needed to be included to improve the match with experimental data. The results of this thesis provide insight on the soot nucleation process and can be utilized to improve current soot formation models.