Outstanding properties emerge at the interfaces of heterogeneous materials, so that their engineering offers promising prospects for achieving novel functional structures. The design and realization of highly-controlled interfaces require reliable characterization techniques with high spatial resolution and chemical sensitivity, and X-rays Energy Dispersive Spectroscopy Spectrum Imaging (XEDS-SI) has been used qualitatively for this purpose with ample success. However, the extraction of quantitative features from XEDS-SI by the overall signal integration around the X-rays peaks location tends to be inaccurate for high spatial resolution analysis due to their typically reduced signal to noise ratio. 

This work presents a strategy for an improved and quantitative chemical analysis at the interface of heterostructures based on the processing of XEDS-SI datasets obtained by aberration-corrected Scanning Transmission Electron Microscopy (STEM). The successive XEDS-SI dataset breakdown with decreasing binning sizes is implemented in the SIev software tool, and its application results in the improved detection of X-rays peaks and estimation of local noise levels. This approach supports the actual chemical signal extraction from XEDS-SI with the maximum spatial resolution with respect to signal to noise ratio (SNR) significance limit.

Results of anisotropic XEDS-SI dataset breakdown obtained with aid of SIev software indicate that sub-nm precision, considering a 2σ confidence level, can be routinely attained on the determination of a projected intermixing layer at the interface of heterogeneous materials. Given that the SIev data processing is significantly faster than a high-SNR XEDS-SI dataset acquisition by the use of currently available high efficiency X-rays detectors, the perspective of an accurate and real-time profile across the interfaces of heterogeneous materials via chemical mapping is foreseen.

Figures:

Figure 1: (top) Atomic resolution High Angle Annular Dark Field (HAADF) from a GaAs-AlGaAs heterostructure interface. (mid) Qualitative colour map indicating the Ga-Kα and Al-Kα signal distribution from a XEDS-SI experiment. (bottom) Ga-Kα and Al-Kα chemical signal profiles across the interface and assessment of the intermixing layer length – 1.2 ± 0.1 nm (2σ)

« Back to The 16th European Microscopy Congress 2016

EMC Abstracts - https://emc-proceedings.com/abstract/siev-implementation-of-an-anisotropic-binning-strategy-to-optimize-the-chemical-analysis-of-heterogeneous-interfaces/