A new method to orient samples by STEM in a scanning electron microscope

Abstract

Transmission electron microscopy is a widely used technique for dislocation characterization. To determine, e.g. dislocation Burgers vectors, the specimen must be oriented in a two‐beam condition where only one Bragg reflection is strongly excited [1]. Similar diffraction information can be in principle obtained by scanning transmission electron microscopy (STEM) in a scanning electron microscope at low electron energies E 0 up to 30 keV. Low‐energy STEM has been shown to be a promising technique [2, 3], which is particularly interesting for the investigation of radiation‐sensitive materials. However, diffraction patterns cannot be taken in a scanning electron microscope without additional instrumental attachments. Some groups already succeeded in obtaining electron channeling patterns by rocking the electron beam on a small sample area while recording backscattered electrons [4]. Here we will present an orientation technique by using the STEM detector in a scanning electron microscope. We will describe the underlying principle and present the orientation procedure for a 100 nm 2 specimen area by using the six‐segment high‐angle annular dark‐field (HAADF) and bright‐field (BF) STEM detectors (cf. Fig. 2). All studies were performed with E 0 = 30 keV. An electron transparent cross‐section specimen of a 500 nm InN layer on a Si‐substrate is shown in Fig. 1, which was used as a test object. Diffraction information at low E 0 can be obtained even at moderate sample thickness and high scattering angles as demonstrated in previous work [5]. The procedure described in the following is suited to orient the electron beam precisely along a zone‐axis orientation. The single crystalline sample was prepared roughly along the [210] InN zone‐axis and the main Kikuchi bands are indicated in Fig. 2 by green, black and grey lines. By tilting the sample, the Kikuchi pattern of the inspected sample region moves across the HAADF detector leading to image intensity changes. The intensity of segments A‐C and D‐F is normalized by the intensity of the incident electron beam on the respective HAADF segments to take into account different possible amplification characteristics. The difference of the normalized intensities between segments A‐C and D‐F is plotted in Fig. 3. After a 2.5 ° tilt the intensity difference between A‐C and D‐F is zero and the green Kikuchi band is oriented along the dashed red line “a” in Fig. 2. To reach a [210] zone‐axis orientation the sample was then tilted around the perpendicular tilt axis. By recording a tilt series around the second tilt axis, the intensity differences in the corresponding segments are compared. In analogy to the first tilt series, the intensity difference for opposite HAADF segments becomes zero for an 8° tilt. The black Kikuchi band in Fig. 2 is then aligned along the dashed red line “b”, and the zone‐axis is reached. A final check can be made by comparing the intensity differences between all opposite HAADF segments, which should be zero in all cases. By tilting the sample around one axis, we can scan the intensity across one Kikuchi line by using the BF detector. Fig. 4 shows the intensity of BF images as a function of the tilt angle. Since the width of the Kikuchi band yields information about the Bragg angle, it can be used to determine the lattice parameter. With the half width of the intensity curve in Fig. 4, the Bragg angle of the green Kikuchi band in Fig. 2 was determined to be ~ 1.25° and 0.6° for the black Kikuchi band. Both values agree well with Bragg angles of 1.14° and 0.704° for the (002) and (1‐20) planes. Based on the analysis of the intensity curves, two‐beam conditions for the inspected area can be obtained if the sample is tilted at angles, which are marked by dashed red lines in Fig. 4. This means that the technique can be applied in low‐energy STEM to characterize dislocations in the future.