| Summary: | Molecular nitrogen, N₂, is the most abundant molecule in the terrestrial atmosphere. Its cation N₂⁺ is therefore prevalent in the earth's ionosphere as well as in nitrogen plasmas produced for reasons varying from lightning strikes to combustion. Any model which seeks to describe plasmas in air must contain a description of nitrogen ion chemistry. Despite this, there is a distinct paucity of data describing electron-N₂⁺ interactions and the resultant bound and quasi-bound electronic structure of N₂. The characterisation of these states is essential for describing dissociative recombination which is the main destroyer of molecular ions in a plasma. This thesis aims to alleviate this problem by performing extensive ab initio R-matrix calculations to create a comprehensive map of the highly-excited electronic structure of N₂ which can the be used to perform a dissociative recombination cross-section calculation. Potential energy curves were found by performing resonant and bound state calculations for all singlet and triplet molecular symmetries of N₂ up to l ≤ 4. The use of a dense grid meant that highly-excited electronic states could be found with an unprecedented level of detail. Many of the states were previously unknown. A new fitting method was developed for the characterisation of resonant states using the time-delay method. It was shown that whilst the R-matrix method is not competitive with conventional quantum chemistry techniques for low lying valence states, it is particularly appropriate for highly-excited states, such as Rydberg states. The data gained from these calculations was then used as an input for a multichannel quantum defect theory calculation of a dissociative recombination cross-section. A description is given of how to prepare the data from the R-matrix calculation for input into a multichannel quantum defect theory dissociative recombination cross-section calculation. Cross-sections were found for v=0-3 including three ionic cores. Whilst previous studies of dissociative recombination using R-matrix data required some empirical intervention, the cross-section found in this thesis is completely ab initio and is in good agreement with experiment.
|
|---|
