Simultaneous mapping of EMCD signals and crystal orientations in a transmission electron microscope

Visualizing subatomic orbital and spin moments using a scanning transmission electron microscope

Magnetism originates from the spin and orbital angular momenta of electrons and their coupling. These interactions occur at subatomic scales and a comprehensive understanding of such phenomena relies on characterization techniques capable of probing the spin and orbital moments at atomic resolution. Although electron energy loss magnetic chiral dichroism has previously enabled the detection of magnetic moments at atomic scales, it was limited to a chromatic-aberration-corrected transmission electron microscope. Although possible, the detection of atomic-scale electron energy loss magnetic chiral dichroism in a scanning transmission electron microscope remains elusive due to challenges associated with convergent beam setups. Here we demonstrate the detection of atomic-scale electron energy loss magnetic chiral dichroism signals in a probe-corrected scanning transmission electron microscope. We not only determine the orbital-to-spin moments ratio for individual atomic planes of an iron crystal but also reveal its local variations at subatomic scales. These findings open the possibility of resolving magnetism down to the orbital level in future studies.

Quantitative magnetic mapping in TEM through accurate 2D thickness determination

Off-axis Electron Holography and Electron Magnetic Circular Dichroism are powerful Transmission Electron Microscopy (TEM) techniques capable of mapping magnetic information with near-atomic spatial resolution. However, the magnetic signals obtained is semi-quantitative due to factors such as thickness variations and local crystallographic changes. Precise determination of spatial thickness variations can make these techniques more quantitative. Electron Energy Loss Spectroscopy (EELS) provides a method to measure thickness variations within a region of interest. The absolute thickness depends on reliable estimates of the inelastic mean free path (λ), which is often unknown for many materials. Alternative techniques, such as Scanning Electron Microscopy (SEM) and Convergent Beam Electron Diffraction (CBED), either lack spatial resolution in thickness mapping or are accurate only within a limited thickness range. Here, we present a straightforward approach to precisely determine the inelastic mean free path (λ), enabling accurate thickness measurements from EELS maps. We compare these thickness measurements with CBED- and SEM-based methods, identifying discrepancies, particularly in thinner samples (<100nm). Finally, we demonstrate how this calibrated thickness measurement can provide quantitative magnetic maps using TEM-based magnetic measurements.

Experimental evidence of magnetism in a 2D electron gas at the CoO/Co3O4 interface by employing EMCD

In the present study we investigate the CoO/Co3O4 interface in order to determine its intriguing magnetic behavior, which can be utilized for tailoring magnetic properties, enabling spin transport, enhancing magnetic coupling, tuning device functionalities, and realizing miniaturized magnetic devices for various technological applications. We decipher the magnetic properties of the CoO/Co3O4 interface from first principles calculations using Wien2k and probe them experimentally by employing electron energy-loss magnetic chiral dichroism (EMCD), which is an electron-energy loss spectrometry (EELS) based technique in the transmission electron microscope (TEM). Both, theory and experiment, are in perfect agreement and result in a ferromagnetic 2D-electron gas of 5Å thickness directly at the interface.

Noise-dependent bias in quantitative STEM-EMCD experiments revealed by bootstrapping

Electron magnetic circular dichroism (EMCD) is a powerful technique for estimating element-specific magnetic moments of materials on nanoscale with the potential to reach atomic resolution in transmission electron microscopes. However, the fundamentally weak EMCD signal strength complicates quantification of magnetic moments, as this requires very high precision, especially in the denominator of the sum rules. Here, we employ a statistical resampling technique known as bootstrapping to an experimental EMCD dataset to produce an empirical estimate of the noise-dependent error distribution resulting from application of EMCD sum rules to bcc iron in a 3-beam orientation. We observe clear experimental evidence that noisy EMCD signals preferentially bias the estimation of magnetic moments, further supporting this with error distributions produced by Monte-Carlo simulations. Finally, we propose guidelines for the recognition and minimization of this bias in the estimation of magnetic moments.

Single scan STEM-EMCD in 3-beam orientation using a quadruple aperture

The need to acquire multiple angle-resolved electron energy loss spectra (EELS) is one of the several critical challenges associated with electron magnetic circular dichroism (EMCD) experiments. If the experiments are performed by scanning a nanometer to atomic-sized electron probe on a specific region of a sample, the precision of the local magnetic information extracted from such data highly depends on the accuracy of the spatial registration between multiple scans. For an EMCD experiment in a 3-beam orientation, this means that the same specimen area must be scanned four times while keeping all the experimental conditions same. This is a non-trivial task as there is a high chance of morphological and chemical modification as well as non-systematic local orientation variations of the crystal between the different scans due to beam damage, contamination and spatial drift. In this work, we employ a custom-made quadruple aperture to acquire the four EELS spectra needed for the EMCD analysis in a single electron beam scan, thus removing the above-mentioned complexities. We demonstrate a quantitative EMCD result for a beam convergence angle corresponding to sub-nm probe size and compare the EMCD results for different detector geometries.