SiON metrology using angular and energy distributions of photoelectrons

Angle-resolved X-ray photoelectron spectroscopy (ARXPS) is a useful tool for non-destructive in-depth analysis of near surface regions. However, the reconstruction of depth profile from ARXPS data is an ill-posed mathematical problem. Thus, the main goal of this work was to develop a new, iterative algorithm based on the least square fitting which allows to solve this problem. The depth profiles were restored by dividing sample in thin virtual box shaped layers each with a different concentration. To extract information on the depth distribution, this algorithm is based on the analysis of the angular peak intensities along with the inelastic background. In addition, the physically trivial constraint of atomic fractions adding up to unity was imposed. The model takes into account the effect of elastic scattering and anisotropy of the photoelectric cross section. To test the algorithm, experimental spectrum for SiON samples on Si substrate were measured with a Thermo Theta Probe electron spectrometer for off-normal emission angles in the range between 25° and 75°. A very good agreement was found between the measured spectra and obtained spectra from the algorithm.