Ph.D. Defense: Final Oral Examination for the Degree of Doctor of Philosophy of Christopher M. Heirwegh

Event Details

  • Date:
  • Time: 1:00 PM
  • Location: MacN 222

Thesis Title:

“Studies of Light Element X-Ray Fundamental Parameters Used in PIXE Analysis"

Abstract:

New measurements of fundamental parameters (FP), used in reference-free particle induced X-ray emission (PIXE) analysis of materials, may be assisted by incorporating both state-of-the-art spectral processing methods and methodologies that have minimal reliance upon other FPs. Under this heading, several light element FPs have been investigated.

A collection of monochromatic spectra have been characterized in terms of the origins and lineshapes of features arising due solely to photon and electron interactions in the silicon detector crystal. Through the assistance of Monte Carlo simulation results, state-of-the-art fit treatments were applied to these spectra using a fit routine that combined both non-linear least squares and manual optimization. The fitted spectra were used to derive a new estimate of the K X-ray fluorescence yield (ωK) of Si. This was done using a geometrical expression that relates ωK to the area ratio of the detector escape to primary peaks taken from the spectra. The final result (0.0504±0.0015) was realized through assistance from new low energy mass-attenuation coefficient (MAC) data.

The accuracy of low-energy (1–2 keV) MACs of light elements was assessed using an approach incorporating PIXE measurements on pure element (Mg, Al and Si) and oxide (MgO, Al2O3 and SiO2) targets. Calculated spectrometer efficiency constants, compared between target pairs (eg, Si vs. SiO2), allow many FPs to cancel. Any non-zero difference in the comparison indicates errors associated with the remaining FPs. A resultant 4 - 6 % discrepancy was attributed to the used of XCOM MACs but this was reduced to 0.5 - 2.5% using FFAST MACs. Additional measurements, performed on silicate micro-probe standards, were analyzed using the same comparative approach. A light element efficiency-constant discrepancy of 7 - 9% was observed and attributed to use of XCOM MACs. This was reduced to 0.5 - 3.5% using FFAST MACs but was reduced further to −0.5 - 2% using a combination of XCOM and FFAST MACs. This result suggested that the combination database was superior.