Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Physics and Astronomy

Committee Member

Endre Takacs, Committee Chair

Committee Member

Chad Sosolik

Committee Member

Marco Ajello

Committee Member

Randall Smith


Astrophysics is a broad and dynamic field that has led to an ever increasing number of incredible discoveries. Just in the past decade or so astrophysicists have detected gravitational waves (and the electromagnetic counterpart) from a neutron star merger, imaged a black hole for the first time, discovered thousands of new planets orbiting stars, and have shown that the expansion of the Universe is accelerating. Many of these discoveries come from new facilities with advanced technologies, an increase in computational capabilities, and creative new analytical techniques. These continued improvements have led to higher quality data that often reveals that our understanding of the processes responsible for the observations is far from complete. It is the field of laboratory astrophysics (experimental and theoretical) that aims to advance our understanding of the underlying processes for more reliable interpretations of astrophysical observations.

With this motivation in mind, this work first describes the electron beam ion trap (EBIT), a facility well suited for systematic atomic studies. The EBIT has a nearly mono-energetic electron beam and allows for the injection of a variety of species, including astrophysically relevant elements such as Fe or Ar. Since ions are present almost everywhere in the Universe, and are responsible for much of the measured emission, it is important to note that the tunable electron beam energy can reach up to about 30 keV and is capable of producing basically all charge states of astrophysically relevant elements. The narrow electron beam energy profile allows the user to select the charge state and to an extent the excited state, and is well suited for systematic studies. The EBIT contains a series of electrodes used to manipulate the electron beam and electrostatically trap the ions. The space charge of the electron beam and shape of the trapping electrodes work to radially trap ions. Observation ports are located radially around the trap and are oriented perpendicular to the direction of the electron beam.

The non-thermal uni-directional electron beam interacts with stationary ions in the trap. This setup leads to non-statistically populated magnetic sublevels that produce polarized and anisotropic emission, and provides a unique opportunity to study magnetic sublevels which are typically inaccessible in spectroscopic observations. In the second part of this work we take advantage of this capability of the EBIT and report the measurement of the linear polarization of He-like and Li-like Ar transitions. Measurements were taken with two Johann-type crystal spectrometers in different orientations corresponding to the dispersion plane parallel and perpendicular to the electron beam direction. The Li-like transitions result from the resonant dielectronic recombination process while the He-like transitions are produced from electron impact excitation. Our results show a strong positive polarization of the w, j, k, and q transitions (in notation of Gabriel (1972)), and a negative polarization of the a, x, y, and z lines.

Since the polarization depends on the magnetic sublevel specific direct excitation or dielectronic capture cross-sections, our results can be used to benchmark different methods used to calculate these cross-sections. In this work we compare measurements with polarization values calculated using the density matrix formalism. For dielectronic recombination, the Flexible Atomic Code (FAC) (Gu 2008) was used to produce the atomic data (Qd values, autoionization energies, and cross-sections) required to calculate the polarization and produce the synthetic spectra. Since measurements were taken at the resonance energy, cascade effects were ignored. For transitions resulting from direct excitation the collisional-radiative model NOMAD (Ralchenko &

Maron 2001) was used to solve the system of steady-state rate equations for the magnetic sublevel populations, and included excitation up to n = 5. For both direct excitation and dielectronic recombination the theoretical predictions agree well with measured values.

The final part of this work was motivated by an exciting 2014 study (Bulbul et al. 2014) that reported a possible dark matter signature at 3.55 keV - 3.57 keV in the stacked spectra of galaxy clusters. To help rule out possible atomic origins suggested by the authors, we measured Ar emission from 1s^(2)2l-1s2l3l"² satellite transitions near 3.6 keV x-ray energy. X-rays were measured simultaneously with a high count-rate, high-purity Ge detector and a high energy-resolution Johann-type crystal spectrometer. The collisional-radiative model NOMAD was used to create synthetic spectra for comparison with both our EBIT measurements and with spectra produced with the AtomDB database (Foster et al. 2012) and the Astrophysical Plasma Emission Code (APEC) (Smith et al. 2001) used in the 2014 work. Excellent agreement was found between the NOMAD and EBIT spectra at each electron beam energy, providing a high level of confidence in the atomic data used. Comparison of the NOMAD and APEC spectra revealed a number of missing lines at 3.56 keV, 3.62 keV, 3.64 keV, and 3.66 keV in the APEC spectra. These features are primarily due to Be-like Ar DR data missing in the database. At an electron temperature of Te = 1.72 keV, the inclusion of 1s2l2l'2l'' and 1s2l2l'3l'' data in AtomDB increased the total flux in the 3.5 keV to 3.66 keV energy band by a factor of 2. While important, this extra emission is not enough to fully explain the unidentified line found in the galaxy cluster spectra (Gall et al. 2019) leaving the possibility open for dark matter related origin.