Electron channelling contrast in the scanning electron microscope (SEM).
Including electron channelling pattern (ECP) and electron channelling contrast imaging (ECCI).
Angus Wilkinson (corresponding)
Electron channelling patterns (ECPs) and electron channelling contrast images (ECCIs) were conceived in the 1960s. Since then there have been several periods of research activity, over the intervening decades, trying to improve and exploit the methods. In the current generation of FEG-SEMs this work has reached fruition as ECCI is capable of routinely imaging lattice defect structures that are so important in controlling properties in structural and functional materials alike.
ECP show the variation of electron signal (typically the back scattered electron intensity) with incident beam direction. These were first reported by Coates (1967) who had been observing large single crystalline semiconductor samples and noted bands of bright contrast linked to the sample alignment rather than position. Coates established many of the properties of these Kikuchi-like reflections, but a much more complete theoretical description was given by Booker et al (1967) using the Bloch wave based dynamical diffraction theory that had recently been used to explain many effects in transmission electron microscopy. There had been good discussion between the two groups prior to publication as a pair of linked papers. It was clear to the Oxford group that the effect implied that much more might be done and predicted that ‘It is only necessary to orientate the crystal at the Bragg position and the local bending of the crystallographic planes where the dislocations emerge at the surface should produce the necessary contrast.’ This accurately describes the ECCI technique but it has taken decades to realise the technique reliably. The notes below attempt to pick-out some of the key contributions along the way.
Coates (1967) – first report of ECP in the SEM. Angular variation of incident beam direction affected by beam scanning across large single crystal samples.
Booker et al (1967) – Theoretical explanation of Coates observations using Bloch wave description. Great foresight in predicting that channelling contrast would allow grain, sub-grain, and lattice defect (dislocations, stacking faults etc) to be imaged in bulk samples in the SEM.
Joy, Newbury & Davidson (1982) – A review of ECP covering physics of pattern formation, electron optics and some applications. The comparison of different strategies for forming ECP with good spatial resolution and angular acuity is a particular highlight and may be of interest to those trying to implement ECP on modern microscopes.
van Essen & Schulson (1969) and van Essen, Schulson & Donaghay (1970) – description of the most successful method for generating selected area electron channelling patterns (SA-ECP). This uses the deflection-focus strategy in which the lower scan coils are switched off and the focusing action of the objective brings the beam back to a focal point on axis after deflection by the upper scan coils.
Clarke & Howie (1971), and Spencer & Humphreys (1980) – two beam dynamical diffraction simulations of ECP contrast profiles and dislocation and stacking fault ECCI intensity profiles. Image contrast forms, along with necessary control of sample alignment and requirements for beam diameter and divergence are all given. The conclusion is that a high brightness instrument is required to realise ECCI of dislocations in practice and that FEG-SEMs might just be able to do this.
Morin et al (1979) and Morin et al (1979) – ground breaking experiments demonstrating that good images of individual dislocation lines and stacking faults near the surface of bulk samples can be obtained in a FEG-SEM. The instrument ran at 45 keV for increased brightness and a concentric spherical grid retarding-field filter used to detect low-loss electrons to increase contrast and signal to noise ratio. SA-ECP were used to control diffraction conditions and dislocation in Si were imaged under a variety of reflections.
Czernuszka et al (1990) – returned to the ECCI method after a significant pause in research in this area. They reported experimental imaging of dislocations using a VG STEM in reflection mode at high energy (100 keV). They used a tilted sample geometry and with a BSE detector positioned at low take-off angle. Images were obtained without recourse to a complex retarding field detector.
Wilkinson et al (1993) and Wilkinson & Hirsch (1995) – Oxford group continue with tilted sample and BSE detector positioned at low take-off angle. Limited work on individual dislocation lines. Clear demonstration of ECCI of overlapping strain fields from groups of dislocations at relatively deep interfaces between epitaxial layers and substrates. Theoretical simulations show the importance of surface relaxation effects for correct characterisation of such defects.
Ahmed et al (1997), Ahmed et al (1999), and Ahmed et al (2001) – a series of papers using ECCI in a SEM with LaB6 source to image dislocation substructure in Cu single crystals after cyclic deformation. Striking images of self-organised patterning in the dislocation structures.
Crimp & Simkin (1999), and Crimp et al (2001) – showed that a fairly conventional geometry of near normal incidence and an annular BSE detector at the pole piece could be used to image dislocations. The work showed the advantage of higher currents, and better signal to noise achievable in a SEM with a Schottky thermal FEG. Demonstration that the g.b=g.bxu=0 invisibility criterion works for ECCI and can be used to determine dislocation Burgers vectors.
[Crimp became interested in ECCI while on sabbatical in Oxford and implemented his own activity on returning to MSU.]