edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 25 NO. 3 12 FOUR-DIMENSIONAL SCANNING TRANSMISSION ELECTRON MICROSCOPY: PART I: IMAGING, STRAIN MAPPING, AND DEFECT DETECTION Aaron C. Johnston-Peck and Andrew A. Herzing National Institute of Standards and Technology, Gaithersburg, Maryland aaron.johnston-peck@nist.gov EDFAAO (2023) 3:12-22 1537-0755/$19.00 ©ASM International® INTRODUCTION Scanning transmission electron microscopy (STEM) has become an indispensable tool for materials characterization due to its ability to elucidate structure and composition with nanometer scale or better spatial resolution. In the scanning transmission electron microscope, electromagnetic lenses form an electron probe that is scanned across a sample (Fig. 1a). As the electron probe interacts with the sample, a distribution of forward scattered electrons (or diffraction pattern) results (Fig. 1b). Within this diffraction pattern, several features can be found, all of which originate from different scattering mechanisms. For crystalline samples, the features that dominate these patterns include sharp, relatively intense discs or spots arising from zero order Laue zone diffraction or Bragg diffraction; Kikuchi bands arising from multiple scattering; or a broad, slowly changing background that is due to several mechanisms including thermal diffuse scattering. Because all these features carry information about the local nature of the sample being probed by the electron beam, electron diffraction patterns are a rich source of information for high spatial resolution materials characterization. However, many traditional imaging modes of STEM and TEM isolate a subregion in the diffraction plane and form an image using only those particular electrons. For example, electrons detected at high scattering angles produce contrast that displays an atomic number dependence, while forming an image using a single diffraction spot produces contrast sensitive to variations in lattice parameter (i.e., strain). In these examples, the images are formed by utilizing a limited fraction of the information generated from the specimen-electron interaction. A paradigm shift would be to detect and record all the scattering information generated and subsequently analyze that dataset to build images that reveal both composition and strain. A new suite of analytical techniques known collectively as fourdimensional scanning transmission electron microscopy (4D-STEM) is that paradigm shift. Keep in mind, 4D-STEM offers much more than the ability to generate “traditional” images. It provides the framework to produce two dimensional (as well as three dimensional) representations of strain, electromagnetic fields, crystallographic phase, and many other material characteristics. This article will introduce the basics of the technique and some areas of application with an emphasis on semiconductor materials. 4D-STEM is a spatially resolved electron diffraction technique that records the electron scattering distribution at each sampling point. The diffraction pattern is projected onto a two-dimensional pixelated recording device providing intensity information as a function of angle (kx, ky). Electromagnetic deflectors scan the electron probe over the sample in a two-dimensional grid of positions (rx, ry), while the two-dimensional recording device is synchronized with the scan to record the diffraction pattern at each position resulting in a four-dimensional dataset (rx, ry, kx, ky) (Fig. 1c). Taking a step back, microbeam electron diffraction, nanobeam electron diffraction (NBED), and convergent beam electron diffraction (CBED) have been used to describe methods where electron diffraction patterns are generated by illuminating a small sample area, but not necessarily coupled with spatially resolved sampling. While reports of spatially resolved diffraction experiments can be found in the literature from as early
RkJQdWJsaXNoZXIy MTMyMzg5NA==