Simulation of Forescattered Electron Channeling Contrast Imaging of Threading Dislocations Penetrating SiC Surfaces

ABSTRACTThe interpretation of ECCI images in the forescattered geometry presents a more complex diffraction configuration than that encountered in the backscattered geometry. Determining the Kikuchi line that is the primary source of image intensity often requires more than simple inspection of the electron-channeling pattern. This problem can be addressed, however, by comparing recorded ECCI images of threading screw dislocations in 4H-SiC with simulated images. An ECCI image of this dislocation is found to give the orientation of the dominant Kikuchi line, greatly simplifying the determination of the diffraction simulation. In addition, computed images of threading screw dislocations in 4H-SiC were found to exhibit channeling contrast essentially identical to that obtained experimentally by ECCI and allowing determination of the dislocation Burgers vector.

Inelastic Scattering Components in the Si(111)-(7×7) RHEED Pattern by the Energy Filtering Method

Energy loss spectra for several parts of the Si(111)-(7 × 7) RHEED pattern, the (0 0) specular spot, the (3/7 3/7) superspot in the zeroth Laue zone, the (0 1) fundamental spot in the first Laue zone, and the Kikuchi line and background have been measured by the recently developed retarding type energy filter in the condition of the [Formula: see text] azimuth with a 10 kV incident electron beam. It was found that there are some differences in their spectra. Energy loss spectra of the (0 0) and (3/7 3/7) spots show surface plasmon loss peaks of silicon dominantly, and the spectrum of the (0 1) spot shows the same but includes weak bulk plasmon peaks. The spectrum of the Kikuchi line mainly shows bulk plasmon peaks and that of the background has no distinct structure in the profile. Glancing angle dependences of their profiles were also measured and discussed. The experimental data show that there is a relation between the quasielastic component of the diffraction beam and the pass length of the electron beam in a vacuum region near the surface where the electron interacts with the surface plasmon. The quasielastic component of the diffraction beam decreases as the incident glancing angle and/or takeoff angle become grazing.

Characterization of Surface Resonance Conditions for Surface Imaging

In order to increase intensity and contrast in the image of a surface, the surface resonance conditions have been widely used to enhance the Bragg reflection for image formation in REM (Reflection Electron Microscopy). However, detailed studies of how the resonance conditions relate to the imaging contrast have not been reported. This paper will concentrate on the general properties of the different resonance conditions, as well as the resulting image contrast.Figure 1 shows a series of RHEED (Reflection High Energy Electron Diffraction) patterns and REM images from the same region of a Pt(l11) surface with the incident electron beam in a direction close to the [112] zone axis at 200 KeV, with a glancing incident angle of about 24 mrad which corresponds to the (555) Bragg reflection condition inside the crystal. For the purpose of convenience in discussion, the four different diffraction conditions shown in figures l(al)-(dl) have been named as D1-D4. With Dl, the specular reflected spot falls in an intersection of a parallel Kikuchi line with a parabola; with D2, the specular reflected spot coincides with an intersection of the Kikuchi lines running parallel to and inclined to the crystal surface; with D3, the specular reflected spot crosses only the parallel Kikuchi line; and with D4, the specular reflected spot intersects only with a parabola. It was found that the diffraction conditions Dl and D2 can not be considered as identical, although the specular reflected spots for both cases are commonly regarded as (555) Bragg reflection in the RHEED pattern. Detailed inspection indicates that for Dl, both the Bragg reflection and the electron surface channelling wave are excited, and for D2, the excitement of simultaneous Bragg reflection occurs closely associated with the properties of three-dimensional dynamical diffraction for a bulk crystal.

On-line measurement of specimen thickness by Convergent Beam Electron Diffraction

Convergent Beam Electron Diffraction (CBED) thickness measurement is the easiest and most accurate way of determining the thickness of crystalline materials. The method was described by Kelly et al. The specimen thickness can be calculated from a few measurements on a recorded diffraction pattern in a matter of minutes (by hand) or seconds (by a computer program).For thickness measurement a CBED pattern is needed that contains a two-beam diffracting condition, with a dark Kikuchi line going through the centre of the Bright-Field disc and the corresponding bright Kikuchi line through the centre of a Dark-Field disc. Parallel to the bright Kikuchi line, the Dark-Field disc contains a number of fringes (Fig. 1) whose distance from the Kikuchi line varies with specimen thickness. The data needed for a measurement are the electron wavelength, the d-spacing dhkl of the diffraction used, the distance 2θB between the Bright-Field disc and Dark-Field disc in the CBED pattern, and the distances Δθi between the dark thickness fringes and the bright Kikuchi line in the Dark-Field disc (Fig. 2).