Files
Abstract
The interactions among the building blocks of our universe: ions, atoms, and molecules, are characterized mainly by collision and scattering. This dissertation provides investigations of three scattering processes with advanced theoretical methods to facilitate more accurate astrophysical modeling. First is the collision between the proton and the hydrogen atom, the most fundamental species present in the Heliosphere, solar/stellar winds, the interstellar medium, etc. The research emphasizes the slow collision region and utilizes the most accurate method, the quantum molecular-orbital close-coupling (QMOCC) approach, suitable for this energy range. Additionally, the research modifies the QMOCC theory by applying the so-called Electron Translation Factor (ETF). Second is the collision between highly charged iron ions from Fe 24+ to Fe 16+ colliding with various neutral targets aiming to explain emission lines observed in the Cygnus Loop and other high-energy astrophysical environments. The work applies the approximate Multi-Channel Landau-Zener method (MCLZ) due to the very large number of inelastic channels. The calculated results are subsequently used in the X-ray cascade model to generate the theoretical X-ray spectrum. Finally, we explore the application of a data-driven, machine learning method to extend SiO-H2 rovibrational collision data. We adopt a novel approach to predict the difference between approximate data and accurate data. This approach has successfully improved the collision data by adding the predicted difference to the approximate data, and it suggests that this approach can be useful in scenarios where abundant approximate data exists, while accurate data are limited, due to their computational expense.