2007 Volume 48 Issue 10 Pages 2580-2584
We propose novel methods of depth-resolved EELS (DREELS) and chemical state mapping, and the techniques were applied to the cross-sectional TEM (XTEM) sample of N+ implanted TiO2 catalyst. DREELS is realized by applying the Pixon deconvolution to a conventional energy-filtering TEM (EFTEM)-based spatially resolved EELS (EFTEM-SREELS), demonstrated by Kimoto et al. [J. Electron Microsc. 46 (1997) 369–374.]. And a self-modeling curve resolution (SMCR) technique in multivariate analysis enabled chemical state mapping from EFTEM-based spectrum imaging (EFTEM-SI) data sets. The methods successfully extracted the depth dependence of the N-K ELNES and separately displayed the spatial distributions of the constituent chemical states, whose spectral features were overlapped.