ABSTRACT
Ronchigrams observed in an aberration-corrected STEM exhibit elaborate interference fringes (Figure 1 ). The angular width and clarity of the region where inference fringes can be seen is much larger than in uncorrected machines. What can we do with these patterns; do they hold any information which we can't obtain from the ADF or conventional TEM image? If we observe a thick region of amorphous material and vary defocus, we see contracting and expanding speckle. Regions near the edge of the pattern change in appearance dramatically. We can also discern different rates of movement (expansion and contraction) as a function of a smooth change of defocus: this is because each level in the specimen is at a different defocus, so that a variety of different apparent local magnifications occur simultaneously. Even when the probe is focussed accurately on the centre of an amorphous specimen, the central disc has elaborate structure, because specimen layers above and below the plane of focus are significantly out of focus. In other words, when their angular range is expanded, Ronchigrams are rich in z-direction data. Indeed, in his original paper on holography, Gabor modelled this exact scattering geometry with light and was surprised to observe 'It is a striking property of these diagrams that they constitute records of threedimensional as well as plane objects. One plane after another of extended objects can be observed in the microscope .. .' [1].
