Automated background subtraction technique for electron energy-loss spectroscopy and application to semiconductor...

Electron energy-loss spectroscopy (EELS) has become a standard tool for identification and sometimes also quantification of elements in materials science. This is important for understanding the chemical and/or structural composition of processed materials. In EELS, the background is often modelled using an inverse power-law function. Core-loss ionization edges are superimposed on top of the dominating background, making it difficult to quantify their intensities. The inverse power-law has to be modelled for each pre-edge region of the ionization edges in the spectrum individually rather than for the entire spectrum. To achieve this, the prerequisite is that one knows all core losses possibly present. The aim of this study is to automatically detect core-loss edges, model the background and extract quantitative elemental maps and profiles of EELS, based on several EELS spectrum images (EELS SI) without any prior knowledge of the material. The algorithm provides elemental maps and concentration profiles by making smart decisions in selecting pre-edge regions and integration ranges. The results of the quantification for a semiconductor thin film heterostructure show high chemical sensitivity, reasonable group III/V intensity ratios but also quantification issues when narrow integration windows are used without deconvolution.