JGIXA — A software package for the calculation and fitting of grazing incidence X-ray fluorescence and X-ray reflectivity data for the characterization of nanometer-layers and ultra-shallow-implants

Grazing incidence XRF (GIXRF) is a very surface sensitive, nondestructive analytical tool making use of the phenomenon of total external reflection of X-rays on smooth polished surfaces. In recent years the method experienced a revival, being a powerful tool for process analysis and control in the fabrication of semiconductor based devices. Due to the downscaling of the process size for semiconductor devices, junction depths as well as layer thicknesses are reduced to a few nanometers, i.e. the length scale where GIXRF is highly sensitive. GIXRF measures the X-ray fluorescence induced by an X-ray beam incident under varying grazing angles and results in angle dependent intensity curves. These curves are correlated to the layer thickness, depth distribution and mass density of the elements in the sample. But the evaluation of these measurements is ambiguous with regard to the exact distribution function for the implants as well as for the thickness and density of nanometer-thin layers. In order to overcome this ambiguity, GIXRF can be combined with X-ray reflectometry (XRR). This is straightforward, as both techniques use similar measurement procedures and the same fundamental physical principles can be used for a combined data evaluation strategy. Such a combined analysis removes ambiguities in the determined physical properties of the studied sample and, being a correlative spectroscopic method, also significantly reduces experimental uncertainties of the individual techniques. In this paper we report our approach to a correlative data analysis, based on a concurrent calculation and fitting of simultaneously recorded GIXRF and XRR data. Based on this approach we developed JGIXA (Java Grazing Incidence X-ray Analysis), a multi-platform software package equipped with a user-friendly graphic user interface (GUI) and offering various optimization algorithms. Software and data evaluation approach were benchmarked by characterizing metal and metal oxide layers on Silicon as well as Arsenic implants in Silicon. The results of the different optimization algorithms have been compared to test the convergence of the algorithms. Finally, simulations for Iron nanoparticles on bulk Silicon and on a W/C multilayer are presented, using the assumption of an unaltered X-ray Standing Wave above the surface