Advanced orientation mapping using precession 4D STEM
 

Advanced orientation mapping using precession 4D STEM  

The goal of this project is to optimize the orientation mapping technique using four-dimensional scanning transmission electron microscopy (4D STEM) in conjunction with precession electron diffraction (PED). The development of complementary metal oxide semiconductor (CMOS)-based cameras has revolutionized the capabilities in data acquisition due to their high sensitivity and fast read out speeds. While scanning an almost parallel, nanometer sized electron probe across the sample, it is now possible to acquire high quality diffraction patterns at each beam position. This produces a complex 4D dataset, where local crystal symmetries, lattice strain and crystal orientation are encoded in the 2D diffraction patterns obtained for each point of the 2D raster grid. The high image quality of the diffraction patterns significantly improves the reliability to determine lattice symmetries and orientations and with this greatly enhances orientation mapping.

Understanding the processing-structure-property relations in nanocrystalline materials has been the main interest of researchers because of their excellent mechanical properties linked with the microstructure. Transmission electron microscopy (TEM) is the most widely used technique because it has the necessary spatial resolution to resolve nanometer sized grains. However, in the case of conventional TEM orientation mapping, orientation and phase identification are often unreliable due to the low sensitivity and ambiguity of spot diffraction patterns. In addition, the acquisition of diffraction data by external charge coupled device (CCD) cameras introduces additional artifacts obstructing orientation and phase identification and requires redundant post image processing such as distortion and inclination corrections.

In this project, we optimize scanning nanobeam diffraction based orientation mapping.  In order to obtain high quality diffraction datasets, we utilize an in-column CMOS camera which is free from distortion and inclination corrections. The large field of view and high dynamic range of the CMOS camera largely enhances even faint details in the diffraction patterns. In combination with PED, the interpretability of each diffraction pattern is improved by increasing the number of diffraction spots and reducing dynamical diffraction effects. Figure 1 shows orientation mapping results of nanoscratched single crystal Si obtained by conventional and optimized process, respectively. Compare to conventional method (Fig. 1b), optimized process minimizes artifacts introduced by the acquisition system (Fig. 1c). Once the optimized scheme is established, we will extend the technique to strain mapping in conjunction with in situ deformation experiments to gain insight into the dynamic strain evolution in nanomaterials. The highly complex datasets of several hundreds of GBs impose another grand challenge that is currently tackled by developing automated routines for data analysis.

Other Interesting Articles

Go to Editor View