Control of magnetic domains and domain walls by themal gradients

Zitieren

Dateien zu dieser Ressource

Prüfsumme: MD5:4029b00a2d82c5d554a1c20b279e772b

STÄRK, Martin, 2015. Control of magnetic domains and domain walls by themal gradients

@phdthesis{Stark2015Contr-32579, title={Control of magnetic domains and domain walls by themal gradients}, year={2015}, author={Stärk, Martin}, address={Konstanz}, school={Universität Konstanz} }

Stärk, Martin 2015 Control of magnetic domains and domain walls by themal gradients The aim of this work is to study heat effects in thin magnetic films to verify the theoretical prediction of domain wall motion towards a heat source. In the theory, a domain wall moves from a cold region towards the hotter part due to energy minimization processes. It is shown that no electron transport is necessary for this effect as it has been observed in insulators. An analytical model describes the effective field that acts on a domain wall as an external applied magnetic field. This is proportional to the temperature gradient and to the exchange stiffness gradient.<br /><br />The theoretical demand for a high temperature gradient is realized in the experiment employing direct laser interference patterning. In the experimental setup a pulsed laser is split into two equivalent beams. These are subsequently recombined on the sample surface at a defined incident angle, polarization and intensity. The result is an interference pattern following a cosine distribution. The period of the pattern is determined by the angle and is varied between 300 nm and 300 µm in our experiment for a wavelength of 532 nm. With the help of this technique, temperature gradients in the range of tenth of K/nm are reached, which should be sufficient for domain wall motion driven by a thermal gradient.<br /><br />To understand the patterning process of direct laser interference patterning in more detail, time-resolved diffraction efficiency measurements are carried out. The results lead to a patterning mechanism which confirms and expands earlier measurements. With the help of time-resolved temperature measurements on an interference pattern, the evolution of the temperature pattern and the temperature gradient are measured in a direct way.<br /><br />The measurement principle is based on the temperature dependence of the refractive index of silicon and interference effects in multilayer sample structures. It is shown that the main part of the inserted heat is transferred through the substrate and the lateral expansion can be neglected.<br /><br />In order to study the effect of temperature on the magnetization of thin films, ferromagnetic [Co/Pd] multilayer systems and ferrimagnetic amorphous Tb<sub>x</sub> Fe<sub>100−x</sub> films are treated with an interference pattern and the magnetic findings are compared to micromagnetic simulations. Here, the patterning period is chosen to be in the range of 50 µm to 70 µm which stretches the important regions laterally.<br />For [Co/Pd] multilayer systems three different magnetic regions are observed after treatment with an interference pattern. In region I, the magnetic material is unaffected and the magnetization is randomly distributed in the so-called AC-demagnetized state. At intensities, which lead to temperatures in the range of the Curie temperature, an abrupt reduction of the domain size is observed, known as thermal demagnetization (region II). At even higher intensities the net out-of-plane magnetization vanishes and the domain pattern is irreversibly destroyed (region III). These findings are supported by micromagnetic simulations indicating that the natural defect concentration and the anisotropy reduction due to diffusion of the multilayer material are the major reasons for the different regions.<br /><br />Contrary to [Co/Pd] multilayer thin films, the ferrimagnetic amorphous material Tb<sub>x</sub>Fe<sub>100−x</sub> exhibits three different regions after laser treatment, but a different explanation is given. In region I and region II the same behavior as in the [Co/Pd] films is shown with the AC-demagnetized region and the thermal demagnetized region, respectively. In addition, region III exhibits the measured magnetic signal as an in-plane magnetization contrast. This is explained from the material point of view as the transition from the amorphous to the polycrystalline phase at a certain temperature, which involves an altered magnetization of the sample. The supporting heat flow simulations confirm the experimental findings.<br /><br />For both materials more complex patterning structures with variable intensities and patterning periods are studied ranging from magnetic thin wires with a width of 120 nm to magnetic bits with an edge length of about 200 nm. In the case of amorphous Tb<sub>x</sub>Fe<sub>100−x</sub> films, a unique way is found to nano-structure the films and to prevent it from oxidation at ambient conditions. As a consequence, the magnetization points in the center of the wire out-of-plane, whereas at the edges an in-plane magnetization is found. In this way, the magnetization of a plain Tb<sub>x</sub>Fe<sub>100−x</sub> thin film is altered subsequently from out-of-plane to in-plane, forming a wire-array on a micro scale with one single interference pulse.<br /><br />Finally, the study of domain wall motion under thermal gradients is described. Therefore, the intensity, the distance and the timescale of the heat source are varied on the two different materials, [Co/Pd] and Tb<sub>x</sub>Fe<sub>100−x</sub>. The experimental setup is capable of a direct observation of the magnetization using the magneto-optical Kerr effect and a simultaneous control of the sample position and the focus of the beam with a submicron precision.<br />The [Co/Pd] multilayer sample is examined as a plain film and as a pre-patterned film using the findings from the material structuring part of this work. Nevertheless, no clear domain wall motion beyond the error bars is found. In contrast, a clear domain wall motion is found in Tb<sub>24</sub>Fe 76</sub> of about 1 µm with a continuous heat source on an unpatterned film. For other stoichiometric compositions of Tb<sub>x</sub>Fe<sub>100−x</sub>, no domain wall motion is observed. This is compared to preliminary theoretical results. It is shown that the exchange stiffness gradient, which is the other important parameter for the effective thermal field acting on a domain wall, differs in ferrimagnetic materials throughout the temperature range. Thus, the exchange stiffness gradient is by a factor of four higher below the compensation point of Tb<sub>x</sub>Fe<sub>100−x</sub> than above. Apart from Tb<sub>24</sub>Fe<sub>76</sub>, the other material compositions exhibit a compensation point below room temperature and therefore, the effect of the exchange stiffness gradient is smaller. In addition, a first distance and intensity dependent study, supported by heat flow simulations, indicates a minor effect of the intensity, but a larger effect of the distance to the domain wall due to the strength of the thermal gradient.<br /><br />For the first time temperature induced domain wall motion in ferrimagnets is observed with a laser beam as a heat source exclusively and no disturbing magnetic fields. 2016-01-14T12:34:38Z 2016-01-14T12:34:38Z eng Stärk, Martin

Dateiabrufe seit 14.01.2016 (Informationen über die Zugriffsstatistik)

Staerk_0-316367.pdf 49

Das Dokument erscheint in:

KOPS Suche


Stöbern

Mein Benutzerkonto