This Is AuburnElectronic Theses and Dissertations

The role of atomic excited states in laboratory plasmas and a study in fine structure diagnostics for far infra-red astrophysical observations




Pearce, A. Jonathan

Type of Degree

PhD Dissertation




This dissertation has a focus on the role of atomic excited states in laboratory and astrophysical plasmas. The emphasis is on systems for which perturbation theory is not expected to produce accurate results. The work includes new electron-impact excitation and ionization data for C + , which is used to produce density and temperature dependent generalized collisional radiative coefficients that can be used for C II spectral diagnostics and C impurity transport. We compare excitation and ionization cross sections with existing calculations and experimental results finding reasonable agreement (within ∼10–15%) for the transitions involving the lower excited states. However, for transitions involving highly excited states (n > 3) the existing electron-impact excitation rate data are from perturbative plane-wave Born calculations and the results from our non-perturbative approach differ greatly. This leads to differences with the currently used ADAS excitation rate coefficients that can be larger than a factor of 10 for these higher n-shells. The issue of final state resolution for C+ excited state ionization is also explored, and evidence is presented that some excited states are unable to ionize directly to the ground term of C 2+ , instead ionizing to the C 2+ metastable term. Additionally, we show that the non-perturbative R-Matrix method can resolve both the initial and the final metastable states in the ionization cross sections. For instance, ionization from the 1s 2 2s 2 2p( 2 P o ) term of C + has a ∼30% peak contribution from the direct ionization to the ground term of C 2+ , ∼20% peak contribution from the direction ionization to the metastable term of C 2+ , and ∼50% contribution from the indirect (excitation-autoionization) processes. Resolution of these final states has significant consequences for impurity transport modeling in laboratory fusion experiments and astrophysical spectroscopy. The dissertation presents an overview of the main differences between these new atomic generalized collisional-radiative rate coefficients and the data currently used in the ADAS database. A study is also presented for neutral N ionization as it introduces a number of scenarios not previously encountered in generalized collisional-radiative modeling. The final-state resolution issues presented for C + are more involved for N and this work is used to outline the methodology for dealing with ionization processes requiring careful consideration of the possible final states, allowing future non-perturbative calculations to fit into the metastable resolved generalized collisional-radiative picture. The remaining topic of this dissertation is the calculation of new fine-structure electron-impact excitation collision strengths for the ground term of Ar 2+ for applications in very low temperature astrophysical plasmas that emit in the far-IR and near UV. The temperature range of interest for this work is 10–1,000K, noting that existing atomic data has not been calculated for temperatures lower than 1,800K. The challenge in diagnosing such low tem- perature plasmas is that the contribution to the cross section from the indirect processes is expected to be very significant, so particular care was given to the resolution of the resonance structure. A progressively finer energy mesh was applied until differences upon successive iterations dropped to less than 0.1%, which resulted in a mesh spacing of 2–40 μeV for the relevant transitions. This fine resolution allowed us to deduce the behavior of the rate coefficients in the 50–100K temperature range that differed significantly from the apparent trends implied by the existing atomic data - by roughly 20%. We evaluate Breit-Pauli R-matrix and Dirac R-matrix calculations, using a comparison of the two datasets to determine the likely uncertainties on the recommended dataset. These uncertainties are typically less than 10% for the temperature range of interest.