A COMPARISON OF ONE-DIMENSIONAL AND 2-DIMENSIONAL SMOOTHING ALGORITHMS FOR AUGER DEPTH PROFILING

The technique of Auger depth profiling is widely use for the chemical analysis of thin film materials. Depth profile data are typically smoothed when they are displayed. This work compares selected smoothing algorithms [one-dimensional (1D) Savitzky-Golay polynomial least squares smooth, two-dimensional (2D) Savitzky-Golay smooth, and factor analysis]. The smoothing routines were judged by comparing the resultant spectral energy resolutions, depth resolutions, and signal-to-noise of spectra and profiles for an anodic oxide film on tantalum metal. The 2D routines were superior to the more conventional 1D smooths for all the criteria considered. The 2D routines also provide the analyst with additional flexibility to trade-off performance in resolution, signal-to-noise, and data acquisition time. The factor analysis procedure was more complex, but gave depth profiling results as good as the best 2D smoothing results.