Detection of magnetic circular dichroism using a transmission electron microscope

Abstract

A material is said to exhibit dichroism if its photon absorption spectrum depends on the polarization of the incident radiation. In the case of X-ray magnetic circular dichroism (XMCD), the absorption cross-section of a ferromagnet or a paramagnet in a magnetic field changes when the helicity of a circularly polarized photon is reversed relative to the magnetization direction. Although similarities between X-ray absorption and electron energy-loss spectroscopy in a transmission electron microscope (TEM) have long been recognized, it has been assumed that extending such equivalence to circular dichroism would require the electron beam in the TEM to be spin-polarized. Recently, it was argued on theoretical grounds that this assumption is probably wrong. Here we report the direct experimental detection of magnetic circular dichroism in a TEM. We compare our measurements of electron energy-loss magnetic chiral dichroism (EMCD) with XMCD spectra obtained from the same specimen that, together with theoretical calculations, show that chiral atomic transitions in a specimen are accessible with inelastic electron scattering under particular scattering conditions. This finding could have important consequences for the study of magnetism on the nanometre and subnanometre scales, as EMCD offers the potential for such spatial resolution down to the nanometre scale while providing depth information--in contrast to X-ray methods, which are mainly surface-sensitive.

Magnetic effects in the paraxial regime of elastic electron scattering

“…Plots of the OAM magnetic signal as function of the collection angle and sample thickness are shown for some example parameters, specified in the captions, in Figs. 4. In all cases one can see a magnetic signal reaching magnitudes of order 10 −5 varying strongly with both thickness and collection angle.…”

Site-specific ionisation edge fine-structure of Rutile in the electron microscope

“…In order to invesitgate the changes in the ionisation edge fine-structure, we use WIEN2k to describe the target's states in the formula above. The MDFF needed in the first simulation package is calculated based on the cross-density of states (XDOS) and the wave functions of the target electron by TELNES.3 subroutines implemented in WIEN2k Schattschneider et al, 2006;Rusz et al, 2007).…”

On the electron vortex beam wavefunction within a crystal

“…This distinction becomes important in, for example, inelastic scattering of atoms at non-zero impact parameters, since the reference frame is usually defined with the atom nucleus at the origin [1][2]. In electron energy loss magnetic circular dichroism (EMCD) measurements this would mean that an off-centred vortex beam would be in a mixed state with respect to the excited atom, rather than being a pure vortex beam of ±ħ angular momentum (note that besides vortex beams there are other methods for measuring EMCD, such as using the crystal as a beam splitter [3]). Indeed Schattschneider et al [4] have shown that in order to detect a strong dichroic signal the vortex probe should be localised at the excited atom.…”