Accurate Nanoscale Crystallography in Real-Space Using Scanning Transmission Electron Microscopy.

Here, we report reproducible and accurate measurement of crystallographic parameters using scanning transmission electron microscopy. This is made possible by removing drift and residual scan distortion. We demonstrate real-space lattice parameter measurements with <0.1% error for complex-layered chalcogenides Bi2Te3, Bi2Se3, and a Bi2Te2.7Se0.3 nanostructured alloy. Pairing the technique with atomic resolution spectroscopy, we connect local structure with chemistry and bonding. Combining these results with density functional theory, we show that the incorporation of Se into Bi2Te3 causes charge redistribution that anomalously increases the van der Waals gap between building blocks of the layered structure. The results show that atomic resolution imaging with electrons can accurately and robustly quantify crystallography at the nanoscale.

Structure of Ultrathin Native Oxides on III-Nitride Surfaces.

When pristine material surfaces are exposed to air, highly reactive broken bonds can promote the formation of surface oxides with structures and properties differing greatly from bulk. Determination of the oxide structure is often elusive through the use of indirect diffraction methods or techniques that probe only the outermost layer. As a result, surface oxides forming on widely used materials, such as group III-nitrides, have not been unambiguously resolved, even though critical properties can depend sensitively on their presence. In this study, aberration corrected scanning transmission electron microscopy reveals directly, and with depth dependence, the structure of ultrathin native oxides that form on AlN and GaN surfaces. Through atomic resolution imaging and spectroscopy, we show that the oxide layers are comprised of tetrahedra-octahedra cation-oxygen units, in an arrangement similar to bulk θ-Al2O3 and ß-Ga2O3. By applying density functional theory, we show that the observed structures are more stable than previously proposed surface oxide models. We place the impact of these observations in the context of key III-nitride growth, device issues, and the recent discovery of two-dimensional nitrides.