Photoelectron holography requires a large data set covering a wide energy range. Changes in experimental conditions and sample surface regeneration during the course of measurement are often unavoidable or necessary, which can result in variations in measured intensity that must be normalized out before data processing. This normalization procedure can introduce a significant error, resulting in image degradation. To eliminate this problem, two methods of intensity self-normalization are introduced, one based on branching ratio measurements and the other based on logarithmic derivative measurements. The As-on-Si(111) system is chosen as a test case for these methods. Both the As and Si core levels are used to reconstruct the three-dimensional atomic structure for the top three atomic layers. The results from these two methods are mutually consistent and in good agreement with other available experimental and theoretical results. This paper also contains an analysis in regard to the use of an inner potential to account for surface refraction. Images are generated without refraction for an estimate of its effect on image quality. Similar analyses are carried out for the proper use of energy and angular window functions in the holographic transform, and for the effect of scattering phase shift that has been a major problem for direct data inversion.
|Number of pages||8|
|Journal||Physical Review B - Condensed Matter and Materials Physics|
|State||Published - 1999|