TEM – Magnetic Materials

  Magnetic Materials


Magnetic materials have attracted the attention and stimulated the imagination of Humankind since the discovery of the attractive properties of lodestone in Ancient China. The potential for practical applications of this “magic” force was soon put in action. The invention of compass in the Middle Ages was just the first of multiple uses of magnetic materials that have contributed to change the world we live in, including the development of electricity in the XIXth century or the development of ultra-high density information storage media. The most recent advances in magnetic materials are related to the emergence of two disciplines intimately related: Nanomagnetism and Spintronics.
New phenomena arise when the physical scale of a magnetic material is reduced down to the nanometric scale, at the threshold between the classical macroscopic world and the quantum phenomena dominating in the range of atomic distances. The magnetic properties of nanoparticles, nanowires or thin films can be drastically different from that of their macroscopic counterparts by the combined effect of the scale reduction and the increasing relevance of surfaces and interfaces. The most remarkable exploitation of the magnetic phenomena at the nanoscale is the development of Spintronics, which relies on the possibility of controlling two main physical properties of the electron, charge and spin, assisted by the use of spin-polarized currents. Spintronics has flourished as one of the main areas of Magnetism for both fundamental research and applications in the fields of information storage and processing, and is intimately related to Nanomagnetism, as most of these phenomena are manifested at the nanoscale in thin films or nanopatterned magnets.

TEM Magnetic Imaging: Lorentz Microscopy and Electron Holography

One of the main issues to understand the deep physical origins of magnetic phenomena at the nanoscale is the accurate determination of the magnetic configurations of nanomagnets, and the analysis of their dynamics as a function of external parameters such as electromagnetic fields or temperature. In this field, Transmission Electron Microscopy provides two different techniques magnetic imaging of magnetic structures with spatial resolution in the range of nanometers: Lorentz microscopy and off-axis Electron Holography, which are intimately related.

Figure 1. Lorentz microscopy in Fresnel mode

The most straightforward magnetic TEM imaging technique is Lorentz microscopy. Although it presents multiple variants, Lorentz microscopy in Fresnel mode is by far the most useful and widely used. Defocused Fresnel images present magnetic contrast (bright/dark lines) at the position of the magnetic domain walls, the strength of this contrast depending on the relative orientation of the magnetization at both sides of the domain wall. The application of the Transport-of-Intensity Equation (TIE) formalism to focal series of Fresnel images enables the retrieval of the magnetic phase shift of the electron wave, which is directly related to the in-plane magnetic induction. Thus, TIE allows the quantitative mapping of the magnetic induction produced by the nanomagnet.

Figure 2. Transport-of-Intensity (TIE) procedure

Off-axis Electron Holography is a more sensitive and quantitative TEM technique sensitive to magnetism. Relying on the interference between a wave interacting with the sample and an unperturbed wave via a negatively biased metallic wire called electron biprism, different strategies can be followed to extract the phase shift of the electron wave induced by the magnetic induction around the sample. The magnetic information can be fully quantitative and enables the mapping of the magnetic induction created by a nano-magnet. An example of holographic reconstruction of the electrostatic and magnetic phases is shown in Figure 3 (Electron Holography).

Figure 3. Electron holography reconstruction of (a) the electrostatic phase shift, and (b) the magnetic phase shift of the electron wave of an ultrathin FEBID Co nanowire. (c) Magnetic flux lines confined inside the nanowire. (d) Profile of the magnetic phase shift perpendicular to the nanowire.

In situ magnetic imaging in TEM

Both techniques require that the sample be in field-free conditions in order to probe the magnetism of the specimen. As the objective lens of a modern microscope applies a magnetic field of approximately 2 T in the sample position, this field would saturate the magnetization of most ferromagnets. Thus, it is required to switch off the objective lens and use an auxiliary focusing lens, the Lorentz lens. It is located immediately below the objective lens and produces a small field in the sample. However, the possibility of applying magnetic fields inside the electron microscope is paramount to probe the magnetic properties of nanomaterials, as it enables the realization of magnetic hysteresis loops in nano-objects to investigate magnetization reversal processes in nanostructures. In this field, we have developed a generalized procedure to apply magnetic field in the sample by slightly exciting the objective lens while tilting the sample. This way, in magnetic specimens with high in-plane shape anisotropy where strong demagnetizing fields makes the out-of-plane component of the external field negligible, we can manipulate the magnetization states of the sample applying controlled in-plane magnetic fields along an arbitrary direction and with amplitudes up to 2000 Oe. This requires a careful calibration of the magnetic field applied by the objective lens as a function of the excitation, which has been carried out with a single tilt TEM holder adapted to carry a Hall probe in the sample position. The result of this calibration can be seen in Figure 4.

Figure 4. Calibration of the magnetic field applied by the SuperTwin objective lens of the FEI Titan Cube.


Figure 5 (Video). Manipulation of domain walls along a cobalt nanowire with the magnetic field applied with the objective lens

The control on the magnetic field we apply to our sample inside the TEM has enabled multiple studies related to the study of magnetic reversal processes of nanostructures (click on the video in Figure 5). For instance, we have carried out a complete characterization of the magnetic nucleation and propagation of domain walls in Focused Electron Beam Induced deposited (FEBID) Co nanowires grown in the Dual Beam facilities of the LMA. In this study we have determined the nucleation and propagation fields of L-shaped nanowires in a wide range of dimensions in order to determine the optimal dimensions for domain wall manipulation. The detection of the nucleation (propagation) fields was carried out using our newly developed procedure for applying magnetic fields and observing the change in contrast of the Fresnel fringes in the curved kink of the nanowires. Surprisingly, a maximum in the nucleation fields is observed in the series of 500-nm-wide nanowires at a thickness of 13 nm, see Figure 7. TIE reconstructions determined that these dimensions correspond to the crossover between the formation of transverse walls (at thickness smaller than 13 nm) and the formation of vortex states (at thickness larger than 13 nm), as shown in Figure 6. Furthermore, the propagation fields for this 500 nm width was nearly independent of the thickness, giving rise to an optimal dimensions for independent, error-free domain wall nucleation and propagation, which could be of great interest for applications of these objects for magnetic data storage or logics.

Figure 6. Domain wall states as a function of width and thickness in FEBID L-shaped cobalt nanowires. Upper row: Lorentz images. Middle row: TIE reconstructions. Lower row: Micromagnetic simulations.


Figure 7. Phase diagram of 500-nm-wide FEBID cobalt nanowires as a function of the domain wall state and the nucleation/propagation fields.

Also in the field of domain wall nucleation and propagation, the high spatial resolution of Lorentz microscopy enabled the analysis of the change of domain wall morphology during depinning domain walls from curved kinks. As shown in Figure 8, in CoFe nanowires, a domain wall originally nucleated as a transverse wall in the kink, transforms into a vortex or a double vortex state while increasing the magnetic field during the depinning process, what confirms that the nature of the domain walls in nano-object does not only depend on the geometry but also on the external bias. This fact determines the functionality of the device because each type of domain wall present different sensitivity and dynamics to the application of magnetic fields or spin-polarized currents.

Figure 8. Underfocus Lorentz images of a CoFe nanowire that show the evolution of a domain walls after depinning. The nucleated transverse wall is transformed in a double vortex wall and propagated either (a) conserving its double vortex configuration, or (b) changing in a single vortex wall. (c) Micromagnetic simulation of the depinning process.


Figure 9. (a) TEM image of the Fe/MgO/FeV/Au tunnel junction. (b) Magnetic induction reconstructions for three different applied magnetic fields. The arrows represent the direction of the magnetization.

In situ magnetic fields can also be applied in holography experiments. A good example is the quantitative analysis of magnetization switching in magnetic heterostructures such as magnetic tunnel junctions (see Figure 9). Electron holography as a function of the magnetic field enables the observation of the (de)coupling of the electrodes as a function of the magnetic field. In fact, as the whole switching process can imaged for the entire nanostructure, the independent determination of the hysteresis loops of each electrode is possible. Furthermore, the perfect control of the applied magnetic field enables to correct for the apparent magnetic induction measured (perpendicular to the electron beam trajectory) with respect to the real magnetization in the sample plane (slightly tilted with respect to the imaging plane). Thus, this correction enables a fully quantitative determination of the magnetization, as can be seen in the Figure 10, which matches perfectly with the simulated values.

Figure 10. (a) Schematics of the relationship between the magnetization in the plane of the simple and the observed magnetization (perpendicular to z axis). (b) Correction of the observed magnetization to estimate the magnetization in the sample plane (parallel to the in-plane magnetic field). (c) Uncorrected and corrected hysteresis loops for the Fe and FeV electrodes of the magnetic tunnel junction. (d) Complete hysteresis loop after correction.


  • Ultra-small functional ferromagnetic nanostructures grown by focused-electron-beam-induced deposition. L. Serrano, R. Córdoba, L.A. Rodríguez, C. Magén, E. Snoeck, C. Gatel, I. Serrano, M. R. Ibarra and J. M. De Teresa ACS Nano 5, 7781-7787 (2011).
  • Optimized cobalt nanowires for domain wall manipulation imaged by in situ Lorentz Microscopy. L. A. Rodríguez, C. Magén, E. Snoeck, L. Serrano-Ramón, C. Gatel, R. Córdoba, E. Martínez-Vecino, L. Torres, J. M. De Teresa and M. R. Ibarra. Applied Physics Letters 102, 022418 (2013).
  • Quantitative in situ magnetization reversal studies in Lorentz microscopy and electron holography. L. A. Rodríguez, C. Magén, E. Snoeck, C. Gatel, L. Marín, L. Serrano-Ramón, J. L. Prieto, M. Muñoz, P. A. Algarabel, L. Morellon, J. M. De Teresa, and M. R. Ibarra. Ultramicroscopy 134, 144-154, (2013).